**Contents**


## **About the Editor**

**Giuseppe Venturella**, Prof., Full Professor of Forest Botany and Mycology at the Department of Agricultural, Food and Forestry Sciences (SAAF) of the University of Palermo, he is also the President of the Italian Medicinal Mushrooms Society (SIFM) and a member of the Mycology Interest Group of the Italian Botanical Society (SBI). He is an Italian representative in the International Society of Medicinal Mushrooms, a member of the Editorial Board of the *International Journal of Medicinal Mushrooms*, as well as the President of the Ninth International Medicinal Mushrooms Conference, held in Palermo in 2017. Finally, he is the author of 108 scientific publications indexed on SCOPUS, and numerous other publications in national and international journals, as well as monographs on mushrooms and truffles.

## *Editorial* **Fungal Diversity in the Mediterranean Area**

#### **Giuseppe Venturella**

Department of Agricultural, Food and Forest Sciences, University of Palermo, Viale delle Scienze, Bldg. 5, I-90128 Palermo, Italy; giuseppe.venturella@unipa.it; Tel.: +39-09123891234

Received: 19 June 2020; Accepted: 19 June 2020; Published: 21 June 2020

**Abstract:** The Special Issue entitled "Fungal Diversity in the Mediterranean Area" aimed at highlighting the role of various organisms in the Mediterranean habitat. The role of fungi at the root and phyllosphere level; the biodiversity in small island territories and the sea; rare forms of fungi never previously found; the commercial, food, and therapeutic value of some ascomycetes and basidiomycetes; the diversity related to fungi associated with galls on plants; and the important role of culture collection for the ex situ conservation of fungal biodiversity are the topics dealt with in this Special Issue.

**Keywords:** fungal diversity; mycorrhiza; Mediterranean forest; medicinal mushroom; bioprospecting; marine fungi; phylogenetics; galls; basidiomycetes; ascomycetes; culture collection

Fungi are extremely heterogeneous organisms characterized by high levels of species diversity and are widespread in all environments. Research on fungal diversity cannot be considered exhaustive, given the continuous discovery of new species and the variability of environments where fungi can be harvested, including the seabed. The fields of application are also varied and range from agriculture, forestry, food, medical, and pharmaceutical sectors. If compared to the central and northern European regions, the Mediterranean environment is a reservoir of continuous discoveries which, in addition to having a taxonomic, environmental, and biogeographical interest, allow researchers to highlight peculiar contents of nutritive elements and uncommon therapeutic applications. This Special Issue includes eight research articles dealing with the fungal biodiversity of the Mediterranean area from various points of view.

Mahmoudi et al. compare samples of roots and rhizospheric soils from arid areas of Tunisia characterized by intensive grazing [1]. The mycorrhizal frequency and the intensity and density of spores varies between plants at the same site and, for each plant, between sites.; Mahmoudi et al. have shown a positive effect of mycorrhizal plants on the microbial activity of the soil. The authors conclude that Arbuscular Mycorrhizal Fungi (AMF) improves soil biological properties, supporting the hypothesis that mycorrhiza and grazing compete for plant photosynthates. Besides, under arid conditions, mycorrhizal symbiosis plays a decisive role concerning soil functionality.

The importance of mycorrhizae is even more evident in the case of species of high historical, gastronomic, and commercial value. *Tuber magnatum* Pico, the most prized truffle in the world, has been studied by Belfiori et al. who examined white truffles from Italy, Hungary, Serbia, Romania, Bulgaria, and Greece and characterized them from a genetic point of view. This study is of fundamental importance for application purposes and to allow the better traceability of white truffles for commercial use and also to prevent the erosion of the biodiversity of white truffles [2].

The biodiversity of macromycetes in Mediterranean forests is the theme of the scientific contributions of Polemis et al. and Gargano et al. In the first article, the authors analyze the fungal diversity of the basidiomycetes associated with *Alnus glutinosa* L. in a restricted environment such as the island of Andros in the Cyclades (Greece). In a long term study, the authors analyze from a morphological, ecological and genetic point of view several macromycetes, of which 21 species are first national records and 68 are reported for the first time from Greek *Alnus glutinosa* forests, including some rare species [3].

Gargano et al. investigated a rare species of albino maitake (*Grifola frondosa* (Dicks.) Gray) collected for the first time in a forest ecosystem of Sicily (southern Italy) [4]. The article highlights the potential application of the albino maitake concerning its nutritional value, particularly high in certain mineral elements and vitamins, and medical value about the ability of its extracts to reduce the production of biofilm by *Staphylococcus aureus* ATCC 43300.

Lazarevi´c and Menkis also highlight how the phyllosphere is expressive of high species diversity. In the case study of the phyllosphere of the endemic forest tree *Pinus heldreichii* H.Christ., a huge number of fungal species were isolated, and mainly constituted Ascomycota [5]. The variability of the fungal community detected at different study sites and altitudes highlights the influence of environmental conditions on the presence/absence of fungal species. There is also a significant correlation between the presence of pathogenic fungi on the leaves, exalted by biotic and abiotic stress factors, and the composition of the fungal community.

The Special Issue also includes an investigation into the diversity of marine fungi by Poli et al. These authors reported the presence of new genera and species isolated from seagrass and algae of the Mediterranean Sea and highlighted how the families Roussoellaceae and Thyridariaceae, until now associated with terrestrial plants, are well represented also in the marine environment [6].

Zimowska et al. contributed to a particular aspect of fungal diversity related to fungi associated with galls on plants of the family Lamiaceae. The results showed full identity with *Botryosphaeria dothidea* (Moug.) Ces. & De Not. of isolates from galls collected from Lamiaceae, while a possible separation from this species should be verified for isolates recovered from *Acacia* in Australia and South Africa [7].

Finally, an interesting contribution to the ex situ conservation of wood decay fungi has been published by Girometta et al. The strains, kept in the MicUNIPV Research Culture Collection of the University of Pavia (Italy), include some species of environmental and medicinal interest closely related to the Mediterranean environment sensu stricto, together with others typical of environments characterized by continental temperate climates [8].

The articles published in this Special Issue reaffirm the importance and role of fungi in different ecosystems. The characterization of fungal biodiversity is of fundamental importance both from an environmental and applicative point of view. Further studies should be conducted in the future to highlight the importance of the in situ and ex situ conservation of fungal diversity for future generations.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


8. Girometta, C.E.; Bernicchia, A.; Baiguera, R.M.; Bracco, F.; Buratti, S.; Cartabia, M.; Picco, A.M.; Savino, E. An Italian Research Culture Collection of Wood Decay Fungi. *Diversity* **2020**, *12*, 58. [CrossRef]

© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Basidiomycetes Associated with** *Alnus glutinosa* **Habitats in Andros Island (Cyclades, Greece)**

#### **Elias Polemis, Vassiliki Fryssouli, Vassileios Daskalopoulos and Georgios I. Zervakis \***

Laboratory of General and Agricultural Microbiology, Agricultural University of Athens, 11855 Athens, Greece; teonanac\_rec1@hotmail.com (E.P.); vfrisouli@gmail.com (V.F.); vassilismks@gmail.com (V.D.)

**\*** Correspondence: zervakis@aua.gr; Tel.: +30-210-5294341

Received: 15 May 2020; Accepted: 7 June 2020; Published: 9 June 2020

**Abstract:** Alluvial forests dominated by black alder (*Alnus glutinosa*) are widespread in Europe along river banks and watercourses forming a habitat of renowned ecological/conservation importance. Despite the considerable interest this habitat has attracted in terms of the associated fungal diversity, very few pertinent data are available from the eastern Mediterranean. Andros island (Aegean Sea, Greece) hosts the southernmost population of *A. glutinosa* in the Balkan Peninsula; such stands have been systematically inventoried for several years in respect to macrofungi. In total, 187 specimens were collected and studied by examining morphoanatomic features and by evaluating (when necessary) the outcome of sequencing the internal transcribed spacer (ITS) region of nuclear ribosomal DNA (nrDNA) to elucidate their identity and obtain an insight into phylogenetic relationships. As a result, 106 species were recorded, 92 are saprotrophic and 14 form ectomycorrhizae (ECM) with alders. Twenty-one species are first national records, while 68 other species are reported for the first time from this habitat in Greece. Several findings of particular interest due to their rarity, ecological preferences and/or taxonomic status are presented in detail and discussed, e.g., six *Alnicola* taxa, *Cortinarius americanus*, *Lactarius obscuratus*, *Paxillus olivellus* and *Russula pumila* (among the ECMs), and the saprotrophs *Entoloma uranochroum*, *Gymnopilus arenophilus*, *Hyphoderma nemorale*, *Lepiota ochraceofulva*, *Phanerochaete livescens* and *Psathyrella hellebosensis*.

**Keywords:** macrofungi; Basidiomycota; mushroom diversity; ectomycorrhiza; saprotroph; alder; Aegean Sea; Mediterranean; *Alnicola*

#### **1. Introduction**

Alluvial forests with *Alnus glutinosa* Gaertn. and *Fraxinus excelsior* L. (priority habitat 91E0\*; Annex I, Directive 92/43/EEC) are distributed throughout Europe, but they are generally rare and threatened since only remnants exist, mainly in central and northern Europe [1]. Alder stands are considerably less frequent in the Mediterranean region, where the repercussions of changes in the hydrological cycle caused by global warming and climate destabilization are much more evident [2]. The southernmost limit of the priority habitat 91E0\* in the Balkan Peninsula is located in Andros island (Figure 1), i.e., the northernmost in the Cyclades and situated at a transition zone between continental Greece and other islands of the Aegean Archipelago. From the geomorphological point of view, it is characterized by a remarkably intense relief and by many rivulets and streams of constant flow, which are unique among most of Central and South Aegean islands. *A. glutinosa* trees demonstrate a patchy distribution in Andros, predominantly occurring along the main streams within the Site of Community Importance (SCI) GR4220001 and in altitudes ranging from sea level to as high as 850 m above sea level (a.s.l.), very close to the highest peaks of the island. In many cases black alders are mixed with *Platanus orientalis* L., *Fraxinus ornus* L. and/or *Nerium oleander* L. (in lower altitudes)*,* while they also form pure stands, as it is the case at the estuaries of the Vori stream in NE Andros.

**Figure 1.** Map presenting Natura 2000 sites, which include the priority habitat 91E0\* in continental Europe (in green) and in islands (in blue); Andros island is indicated by the red arrow. Data from https://www.eea.europa.eu/data-and-maps/data/natura-6.

Alder trees are known to form symbiotic relationships with nitrogen-fixing actinomycetes of the genus *Frankia* Brunchorst [3,4], with arbuscular mycorrhizal fungi (AM) of Glomeromycota [5,6] and with various ectomycorrhizal (ECM) fungi of Ascomycota and Basidiomycota [7–9]. European alder stands have been relatively well-studied in terms of both macro- and microfungal communities, and approx. 1000 species of saprotrophic and ECM macrofungi were reported [10–15]. In addition, mycocoenological studies from Europe and North America suggested that ECM fungi of *Alnus* spp. exhibit a remarkably high degree of host specificity compared to other tree species [8,16], while the analysis of both sporophores and ectomycorrhizae evidenced that alders have a low number (<50) of ECM symbionts worldwide [17–19].

Limited knowledge is available on the diversity of fungi associated with alders in Greece, and only preliminary data are reported in the few pertinent publications [20,21]. On the other hand, Andros is the only island of the Aegean Archipelago where a systematic inventory of macrofungi is in progress for more than 20 years. Biotopes characterized by river banks, springs and alluvial forests, where *A. glutinosa* is often the dominant tree species, were forayed in the past and 37 mushroom species were reported from this particular habitat in Andros, including ECM symbionts as well as xylotrophic, litter and/or humus saprotrophs [22–24]. Among the latter, *Entoloma alnicola* Noordel. & Polemis was described asnew species for science and it is still known from the type locality only [25].

Since 2017, mycodiversity studies in alder stands of Andros were intensified in the frame of a LIFE Nature project (LIFE16-NAT\_GR\_000606), which -among others- aims at the conservation and restoration of the priority habitat 91E0\* in the island. Hence, during the last few years, new sites with alder stands were repeatedly forayed (in addition to those previously investigated), and a large number of new collections were made. These, together with previously sampled—but still unidentified—specimens, were subjected to detailed morphoanatomical examination in conjunction with sequencing and phylogenetic analyses (where judged necessary) in order to assess their identity. Moreover, in several occasions, past relevant reports on recorded taxa were revised/re-evaluated according to the latest respective taxonomic and phylogenetic concepts. Hence, this work presents an updated compilation of available data on the diversity of macrofungi in a habitat of significant interest occurring at the limits of its distribution in Europe.

#### **2. Materials and Methods**

#### *2.1. Sampling of Biological Material*

Data presented in this inventory are based on specimens collected from 10 sampling sites covering almost the entire area of *A. glutinosa* distribution in Andros island, which appears mainly within (or marginally out) the SCI GR4220001, extending from sea-level to an altitude of ca. 850 m a.s.l. (Figure 2; Table S1).The biological material examined for the purpose of this work was sampled in 38 forays performed during the last 25 years from late October to April; more than half of those (#23) were conducted in the period from 2017 to 2020. In total, 187 specimens found exclusively under alder trees or directly on their wood, woody residues or leaf-litter were collected, and voucher specimens are deposited in the Fungarium of the Laboratory of General and Agricultural University of Athens (ACAM).

**Figure 2.** Sampling sites (in yellow marking) in the *Alnus glutinosa* habitat and relative position/size of the area under investigation within Andros island (map in upper right corner).

#### *2.2. Morpho-Anatomical Features in Basidiomes*

The morphological study included in situ recording of macroscopic features of taxonomic interest, while ex-situ examination involved observations of morphoanatomical characters in dried specimens. Sections were mounted and observed in KOH 3–5% (*w*/*v*), in Melzer's reagent, in cotton-blue, in cresyl-blue and in sulfovaniline solution. Observations were performed with the use of a Zeiss AxioImager A2 microscope under bright field and differential interference contrast (DIC); microphotographs were taken with the aid of a mounted digital camera (Axiocam). For all examined specimens a minimum of 30 mature basidiospores were measured and the resulting measurements as well as additional observations of other essential microscopical features (hymenial cystidia, pileipellis etc.) were used for determination of the species examined in accordance to pertinent identification keys and monographs (e.g., [26–36]).

#### *2.3. DNA Extraction, Amplification and Sequencing*

When deemed necessary, DNA sequencing and phylogenetic analyses were performed. Total genomic DNA was obtained from dried basidiomes and DNA extraction was performed through the use of the Nucleospin Plant II DNA kit (Macherey and Nagel, Düren, Germany) by following the manufacturer's protocol. The internal transcribed spacer (ITS; ITS1, 5.8S, ITS2) region within the nuclear ribosomal RNA gene cluster was examined by using the primers ITS1/ITS4 [37]. Polymerase chain reactions (PCR) were performed in 50 μL containing 50 ng DNA template, 0.25 μM of each primer, 0.2 mM of each dNTP, 1 × HiFi Buffer (Takara BIO INC., Shiga, Japan) and 1 U HiFi Taq DNA polymerase (Takara BIO INC., Shiga, Japan). PCR reactions were performed as follows: 94 ◦C for 5 min, followed by 35 cycles of 94 ◦C for 30 s, 50 ◦C for 30 s and 72 ◦C for 1 min, and a final extension at 72 ◦C for 10 min. PCR products were run in 1% agarose gels and purified using Invitrogen PureLink kit (Thermo Fisher Scientific, Seoul, S. Korea), and were submitted for sequencing to CeMIA SA (Larissa, Greece). The same PCR primers were used for sequencing. Chromatograms were checked with the aid of BioEdit v. 7.2.5 software [38]. Then sequences were examined against GenBank built-in search tools for obtaining information which could confer at identifying the material under study. A total of 61 validated sequences generated in this work were deposited in GenBank and the accession numbers MT458502 to MT458562 were obtained.

#### *2.4. Phylogenetic Analysis of Sequence Data*

A total of 42, 29 and 22 ITS sequences corresponding to selected species of the genera *Alnicola* Kühner (and *Naucoria* (Fr.) P. Kumm.), *Lactarius* Pers. and *Paxillus* Fr. (including 12, 5 and 4 sequences generated in this work), respectively, were subjected to phylogenetic analysis. In addition, species of the same or other genera were used as outgroups in each case. Multiple sequence alignment of each ITS rDNA dataset was conducted using the Q-INS-I algorithm as implemented in the online version of MAFFT v. 7 [39]. Alignments were reviewed, manually adjusted at misaligned sites and trimmed at the same position through MEGA X [40] before being used for further analysis.

Phylogenetic relationships of taxa for each alignment were inferred by using maximum likelihood (ML) and Bayesian inference (BI) through the CIPRES web portal (www.phylo.org; Miller et al. 2010). ML analyses were conducted by RAxML BlackBox online server (http://phylobench.vital-it.ch/raxmlbb/) [41] using default parameters and calculating bootstrap statistics according to the program recommendations for the best-scoring ML tree. BI analyses were performed by MrBayes v. 3.2.1 [42]. The best-fit substitution model for each dataset was selected according to the corrected Akaike information criterion (cAIC), as implemented in jModeltest v.2 [43]. The TPM2uf+G, TPM1uf+G and SYM+G models were selected for the *Alnicola*, *Lactarius* and *Paxillus* datasets, respectively. To estimate posterior probabilities, Markov chain Monte Carlo (MCMC) simulation was implemented in two parallel independent runs of four chains, one cold and three heated, with trees sampled every 1000 generations until the standard deviation of split frequencies is below 0.05; the first 25% of trees were omitted as

burn-in. A 50% majority rule consensus tree was built and visualized with iTOL [44]. Clades with ML bootstrap support (MLB) ≥ 65% and Bayesian posterior probability (BPP) ≥ 95% were considered as significantly supported.

#### **3. Results and Discussion**

The study of 187 specimens of macrofungi associated with the *A. glutinosa* priority habitat in Andros led to the identification of 106 species (74 genera) of basidiomycetes. Among them, 14 (13%) are ECM species (Table 1) strictly associated with alders [18,19]. The other 92 (87%) are saprotrophic; 70 (66%) saproxylic and 22 (21%) saprotrophic on soil, humus or leaf-litter (Table 2). Interestingly, 10 ECM and 11 saprotrophic species are first national records, while other 68 are reported for the first time from this habitat in Greece. Identification of specimens to species was performed by examining their morphoanatomic features and by evaluating (when necessary) the outcome of ITS sequencing and phylogenetic analysis; in the latter case, the respective GenBank accession numbers are provided (Tables 1 and 2). Selected findings of particular interest are presented (and discussed) by providing brief descriptions and comments on characters of potentially diagnostic value.




**Table 2.** Saprotrophic basidiomycetes identified during the study: species name, specimen code/collection date, locality, type of substrate and GenBank accession numbers for ITS sequences generated. First national records for Greece are indicated by an asterisk (\*) before the species name.





#### *3.1. The ECM Element*

Among the ECM macrofungi recorded (Table 1), the genus *Alnicola* is represented by six species (Figure 3); five of them form part of the sect. *Alnicola* sensu Moreau [45], which is characterized by urticoid cheilocystidia, and one of the sect. *Submelinoideae* Singer with clavate or capitate cheilocystidia [46].

**Figure 3.** Species of the genus *Alnicola* recorded in Andros alder stands: *A. escharoides* basidiomes, basidiospores and cheilocystidia (**a**–**c**); *A. umbrina* basidiomes, basidiospores and cheilocystidia (**d**–**f**); *A. striatula* basidiomes, basidiospores and cheilocystidia (**g**–**i**); *A. subconspersa* basidiomes, basidiospores and cheilocystidia (**j**–**l**); *A. luteolofibrillosa* basidiomes, basidiospores and cheilocystidia (**m**–**o**); *A. inculta* basidiomes, basidiospores, basidia and cheilocystidia (**p**–**s**). Bars: basidiomes, 1 mm; basidiospores and basidia, 10 μm; cheilocystidia, 20 μm.

Different opinions exist regarding the genus name in pertinent literature since some European authors as well as the Index Fungorum prefer to conserve the name *Naucoria* (Fr.) P. Kumm., whereas Moreau, in his nomenclatural revision, rejected this name in favour of *Alnicola* Kühner [45]; the latter approach is accepted by other European mycologists, the Mycobank, and is also adopted in this work. Moreover, the taxonomy of species of the sect. *Alnicola* remains problematic and, consequently, a phylogenetic analysis was performed to deal with this issue.

The most often found *Alnicola* species in our study were *A. umbrina* (R. Maire) Kühner and *A. escharoides* (Fr.) Romagn., i.e., two of the most common taxa associated with alders in Europe; both constitute new national records for Greece. Particularly *A. escharoides* (syn. *A. citrinella* Moreau & A. de Haan [47]) is distinguished from all other (more or less brownish) species found in Andros by its pale yellowish-buff non striate pileus, the amygdaliform to navicular spores, with prominent ornamentation, measuring 9.9–11.8 × 5.3–5.9 μm, Q = 1.9–2.1 (Figure 3a–c). Following phylogenetic analysis, our specimens are positioned in a distinct group (albeit not adequately supported) together with other sequences from material identified as *A. escharoides* and *A. citrinella* (Figure 4).

**Figure 4.** Phylogeny of *Alnicola* species derived from rDNA ITS sequences through ML analysis. Branches are labelled when MLB > 65% and BPP > 0.95. *Hebeloma* species (*H. louiseae*, *H. pallidolabiatum, H. crustuliniforme*) were used as outgroups. Boxes include sequences from specimens recorded in the *Alnus glutinosa* habitat.

On the other hand, *A. umbrina* (Figure 3d–f) is hereby considered as a species complex following the nomenclatural concept of Moreau [45] and the outcome of the phylogenetic study by Rochet et al. [19]. According to our observations, *A. umbrina* shows a rather large morphological variability with dark brown hygrophanous pilei bearing prominent striations up to their centre when wet, becoming much lighter and indistinctly striate only at margin when dry. Basidiospores are variable in size and shape,

often somewhat elongated fusiform, weakly to moderately verrucose, measuring 10.7–13.6 × 5.2–6.1 μm, Q = 1.9–2.4. Sequences generated in this work clustered together with material identified as *A. umbrina*, *N. scolecina* (Fr.) Quél., *A. striatula* (P.D. Orton) Romagn. and *A. subconspersa* (Kühner ex P.D. Orton) Bon into a group that was not adequately supported (Figure 4). However, the morphological features of specimens identified as *N. scolecina* in Europe are very similar to descriptions of *A. umbrina* [22,33,48,49]. Therefore, *N. scolecina* and *A. umbrina* form part of the same complex and the question whether they constitute different entities or not remains open and in need of further research.

One collection representing another closely related taxon, previously reported as *N. striatula* P.D. Orton (Figure 3g–i) from alder stands in Andros [22], derived from the same site during our recent forays. According to Moreau (2005), *A. striatula* might merely correspond to a pale form of *A. umbrina*, but our morphological studies revealed some noteworthy differences when compared to specimens hereby named *A. umbrina*, i.e., pileus always very prominently striate, smooth and shiny, and (most importantly) significantly smaller basidiospores measuring 8.2–10.0 × 4.5–5.6 μm, Q = 1.7–1.9; these features are in accordance to previous descriptions of *N. striatula* [33,48,50]. As evidenced from our phylogenetic analysis (Figure 4), this particular collection forms part of the *A. umbrina* complex (together with the other two *A. striatula* sequences included in the tree) by using ITS alone; however, since it is morphologically distinct and fits to the widely accepted taxonomic concept of *A. striatula*, we provisionally retain it in this inventory as a separate taxon, until a future multigene approach shows otherwise.

A similar looking species to *A. umbrina*—but less common in Andros—is *A. subconspersa* (Figure 3j–l). The most prominent distinguishing features versus our *A. umbrina* specimens are the non (or very faintly) striate pileus as well as the size and shape of spores, being wider, amygdaliform to navicular and more prominently ornamented, measuring 10.9–12.9 × 6.0–6.8 μm, Q = 1.7–2.0. It is noteworthy that *A. subconspersa* forms a well-supported phylogenetic group including sequences labelled as *A. scolecina* (Fr.) Romagn. (Figure 4), which is indicative of the morphological affinity of these taxa that had apparently led to the development of ambiguous species concepts.

Another collection representing a member of the sect. *Alnicola* was recorded in the alluvial littoral forest of Vori; it corresponds to *A. luteolofibrillosa* Kühner and constitutes the first report of this species in Greece (Figure 3m–o). It is morphologically characterized by non-striate, pale buff, fibrillose to tomentose pilei, with abundant whitish veil remnants on stipe and pileal margin; the respective sequence falls within a highly-supported terminal subgroup corresponding to this species (Figure 4). Lastly, *A. inculta* (Peck) Singer (Figure 3p–s) was recorded only at Zenio (i.e., the site with the highest altitude among those of this study, 850 m) and is reported for the first time in Greece. It forms part of the sect. *Submelinoideae*, and, according to Moreau [45] is conspecific to the taxon widely referred as *N. celluloderma* P.D. Orton, as it is also evidenced by our phylogenetic analysis (Figure 4). Morphologically, this species is easily distinguished from all aforementioned taxa thanks to the clavate to capitate cheilocystidia characterizing members of sect. *Submelinoideae* and the 2-spored basidia.

The most common ECM mushroom in alder stands of Andros belongs to the genus *Paxillus*; it was the first recorded *Alnus*-specific symbiont in the island 25 years ago, and was later repeatedly found in this particular habitat (Figure 5a–d). It was initially identified as *P. rubicundulus* P.D. Orton [22]; however, sequencing of recent collections revealed that it forms part of the newly described taxon *P. olivellus* Moreau P-A, Chaumeton J-P, Gryta H, Jargeat P [51]. Although clearly separated by molecular approaches, *P. olivellus* can be hardly distinguished from *P. rubicundulus* and *P. adelphus* Chaumeton JP, Gryta H, Jargeat P, Moreau P-A on the basis of morphology alone, i.e., only by the olivaceous tinges of the young basidiomes and the basidiospores shape, which are ovoid to ellipsoid in *P. olivellus*, cylindrical in *P. rubicundulus* and short cylindrical in *P. adelphus* [51]. Such features were observed in our specimens since olivaceous tints were always evident in young basidiomes, and spores were ovoid to ellipsoid measuring 6.7–8.1× 4.5–5.2 μm, Q = 1.39–1.66. In addition, ITS sequences from our material originating from various sites in the habitat under study were very similar or identical to those corresponding to *P. olivellus* (including the type), and formed a terminal subgroup with high

support (Figure 6). Therefore, this particular species seems to be the only representative of the genus *Paxillus* in the black alder stands of Andros island.

**Figure 5.** Alder-associated ECM fungi recorded in Andros: *Paxillus olivellus* basidiomes (**a**; bar 1 mm), basidiospores (**b**; bar 10 μm), section of lamella (**c**; bar 20 μm), hymenial cystidium and basidia (**d**; bar 20 μm); *Lactarius obscuratus* basidiomes (**e**, bar 1 mm), basidiospores (**f**, bar 10 μm), pileipellis (**g**, bar 20 μm); *Russula pumila* basidiomes (**h**, bar 1 mm), basidiospores (**i**, bar 10 μm), pileipellis (**j**, bar 20 μm); *Cortinarius americanus* basidiomes (**k**, bar 1 mm); *Inocybe calospora* basidiospores, basidium and pleurocystidium (**l**, bar 10 μm).

**Figure 6.** Phylogeny of *Paxillus* species derived from rDNA ITS sequences through ML analysis. Branches are labelled when MLB > 65%, and BPP > 0.95. *P. cuprinus* and *P. obscurosporus* were used as outgroups. The coloured box includes sequences from specimens recorded in the *Alnus glutinosa* habitat.

*Lactarius obscuratus* (Lasch) Fr. is one the few *Alnus*-specific ECM symbionts of this particular genus; it was found in several inland collection sites dominated by *A. glutinosa*, but not in the alluvial (littoral) forest of Vori (Figure 5e–g). Phylogenetic analysis of our sequences derived from several collections confirmed that they belong to this particular species (Figure 7). However, the respective terminal subgroup in our phylogenetic tree is composed from sequences named either *L. obscuratus* or *L. cyathuliformis* Bon, which is due to the different interpretations existing about this taxon (J. Nuytinck, pers. comm.). In the study of Rochet et al. [19], it is referred as *L. cyathuliformis*, whereas the correct name for the same group is *L. obscuratus* according to Wisitrassameewong et al. [52]. Since the basidiospores average size in our collections (measuring 7.6–8.5 × 6.1–6.3 μm) is in agreement with the concept of *L. obscuratus* (sensu Heilmann-Clausen et al. [31]; according to the same authors, spores of *L. cyathuliformis* have an average size of 8.3–9.9 × 7.0–7.7 μm), we adopt this name for the specimens included in this work. The genus *Russula* Pers. is represented by *R. pumila* Rouzeau & F. Massart (Figure 5h–j) detected in one site only (Vourkoti). Although *R. pumila* is synonymous to *R. alnetorum* Romagn. according to both Index Fungorum and Mycobank, there are different opinions about the synonymy of these two taxa and to the best of our knowledge this issue has not been resolved yet (S. Adamcik, pers. comm.). *R. pumila* is reported to occur mainly in lowlands with *A. glutinosa*, as opposed to *R. alnetorum*, which is mostly recorded in subalpine habitats with *A. viridis* [53,54]; therefore, we adopt the use of the former name.

*Tomentella sublilacina* (Ellis & Holw.) Wakef. and *T. stuposa* (Link) Stalpers were previously recorded in Andros (in the alluvial alder forest of Vori; [20]) and identified on the basis of their morphology. These names are provisionally retained here due to the absence of precise taxonomic information concerning alder-specific *Tomentella* species. It should be noted that the ITS sequences generated in the frame of this work represent phylogenetically distinct taxa corresponding to entities named "aff. *sublilacina*" and "aff. *stuposa*" in previous studies referring to material originating from alder hosts only [18,55].

It is noteworthy that the first ever report of an alniphilous *Cortinarius* species in Greece derives from a single collection of *C. americanus* A.H. Sm. (Figure 5k), which forms part of a small group of species within the subgenus *Telamonia* (Fr.) Trog known to be associated with *A. glutinosa* and *A. incana* in Europe [33,56]. *C. americanus* is characterized by the minute size, pileus not exceeding 2 cm in diameter, with dark violet to blackish colour, and spores smaller than 10 × 6 μm [33,56]; our collection has spores measuring 7.6–8.4 × 4.9–5.6 μm.

Last, *Inocybe calospora* Quél., recorded for the first time in Greece, was collected in Vourkoti only; it is an easily identified species thanks to its unique star-shaped spiny spores (Figure 5). Among all ECM species included in this inventory it is the only one which is not considered to be exclusively associated with alders, and reported from diverse damp deciduous forests of Europe [33,56].

**Figure 7.** Phylogeny of *Lactarius* species derived from rDNA ITS sequences through ML analysis. Branches are labelled when MLB > 65%, and BPP > 0.95. *L. torminosus*, *L. scrobiculatus* and *L. pseudoscrobiculatus* were used as outgroup. The colored box includes sequences from specimens recorded in the *Alnus glutinosa* habitat.

#### *3.2. The Saproxylic Element*

By far the highest number of species recorded in this inventory correspond to white-rot and brown-rot basidiomycetes found on various wood parts of *A. glutinosa* (Table 2). This is quite anticipated since alder trees have a life-span which rarely exceeds 100 years; therefore, they produce large amounts of dead wood. Moreover, in contrast to ECM species, wood-rotting fungi do not show any specificity to alders, with only few exceptions including the common in northern Europe plant-pathogenic polypore *Inonotus radiatus* (Sowerby) P. Karst. [15,57–60], which, however, was not among our findings. Most of the recorded species are wood rotting basidiomycetes that are quite common throughout Europe on deciduous tree species including alders [13,61,62]. The best represented genera of saproxylic fungi were *Hyphoderma* Wallr. (three spp.), *Mycena* (Pers.) Roussel (four spp.), *Pluteus* Fr. (four spp.) and *Psathyrella* (Fr.) Quél. (four spp. on woody residues or buried wood). In total, seven species recorded on dead wood or bark of living alder trees are recorded for the first time in Greece and are presented in more detail below.

Among corticioid basidiomycetes, *Hyphoderma nemorale* K.H. Larss. is a distinct, widely distributed but rare species in Europe [35], which was identified by re-examining an old collection from Vori alluvial forest and further confirmed by ITS sequencing. The presence of thick-walled subicular hyphae and of two types of hymenial cystidia (i.e., short ventricose and subcapitate, and more seldom long tubular with characteristic constrictions, sometimes moniliform, which was the case in our specimen) are the main diagnostic features of this species [63]. *Hyphodermella corrugata* (Fr.) J. Erikss. & Ryvarden (Figure 8a) is fairly common and widespread in Europe [35]. It is easily identified thanks to its characteristic cystidioid hyphal ends appearing in bundles that are heavily incrusted [24]. *Phanerochaete livescens* (P. Karst.) Volobuev & Spirin (Figure 8b) is a species closely related to *Ph. sordida* (P. Karst.) J. Erikss. & Ryvarden which was recently described by using both morphological and phylogenetic criteria [64]. In accordance to the pertinent description, our specimens possessed cystidia with thickened walls to the acute apex, densely covered by crystals, as opposed to the accidentally encrusted, obtuse and thin-walled towards the apex cystidia of *Ph. sordida*. Moreover, the identity of our specimen was confirmed by the respective ITS sequence which was identical to those of *Ph. livescens* as determined by Volobuev et al. [64].

**Figure 8.** Saproxylic species recorded on Alnus glutinosa wood and litter in Andros: Hyphodermella corrugata (**a**) Phanerochaete livescens (**b**) Pluteus podospileus (**c**) Delicatula integrella (**d**) Coprinopsis melanthina (**e**) Gymnopilus arenophilus (**f**) Hydropus floccipes (**g**) Lepiota ochraceofulva (**h**) Lepista ovispora (**i**) Psathyrella hellebosensis (**j**) Entoloma uranochroum (**k**) Bar: 1 mm.

Among the agaricoid wood-inhabiting fungi, the genus *Pluteus* Fr. is hereby represented by four species, of which *P. podospileus* Sacc. & Cub. (Figure 8c) is reported for the first time in Greece. It belongs to the sect. *Celluloderma* Fay. subsection *Mixtini* Sing. ex Sing, and possesses a pileipellis made up of both fusiform and broadly clavate elements. *P. thomsonii* (Berk. & Broome) Dennis is very similar morphologically but it differs in the absence of pleurocystidia and the shape of cheilocystidia, which are characteristically rostrate [65]. The record of *Delicatula integrella* (Pers.) Rat. (Figure 8d) is worth mentioning since it was reported only once before in Greece, in the content of a regional field-guide [66]. *D. integrella* forms whitish-mycenoid mushrooms of minute size with pileus diameter measuring (in our specimens) 0.4–0.6 cm, reduced, almost vein-like lamellae and non-amyloid, amygdaliform-fusoid spores. It is considered widespread and common in Europe and N. America, and grows on decaying wood and wood debris of deciduous trees [34]. *Coprinopsis melanthina* (Fr.) Örstadius & E. Larss. (Figure 8e) is a striking-looking psathyrelloid species, easily identified due to the relatively large basidiomes with wooly to squamulose pileus (measuring 2–6 cm in diam.) and stipe (up to 6.0 × 0.8 cm), absence of pleurocystidia, almost colourless basidiospores, devoid of germ-pore, measuring 9.8–11.8(13.5) × 5.6–6.5 μm in our collections. This is a rather uncommon European species growing on and around rotten stumps in humid deciduous forests [33]; the only other record of this species in Greece derives from Crete (G. Konstandinidis, pers. comm.).

*Gymnopilus arenophilus* A. Ortega & Esteve-Rav. (Figure 8f) was described from continental areas of Spain [67] and from maritime dunes in France, under or near Mediterranean pines, on sandy soil by being attached to wood debris or wood, often burnt or buried in the sand [68]. Two of our collections from the alluvial *A. glutinosa* habitat at Lefka were sequenced and found to correspond to this species. Apparently, no native pines exist in Andros while both specimens were growing on rotten alder stumps, a fact that largely expands the so far known ecological and geographical range of this Mediterranean species. Morphologically, our specimens possessed features that fit well to the taxonomic concept of *G. arenophilus* i.e., smooth to fibrillose pileal surface, bitter taste, ellipsoid to subamygdaliform, moderately verrucose spores, measuring 8–10 × 5.5–6.5 μm, lageniform cheilocysidia often with subcapitate apex, 25–45 × 5–8 μm and absence of pleurocystidia. On the other hand, the size of basidiomes was significantly larger, with pilei up to 10 cm in diam. and a sturdy stipe often thicker than 1 cm. One previous collection of ours also found on rotten alder stump, and originally identified as *G. picreus* (Pers.) P. Karst. [22], is now re-assessed as *G. arenophilus*.

*Hydropus floccipes* (Fr.) Singer (Figure 8g) is a rare mycenoid species generally found to grow on decayed trunks of deciduous trees in damp forests, and is characterized by non-amyloid, subglobose spores, not blackening basidiomes, and typically scabrous stipe with grey-brown spots [33,69]. In addition, our specimens possessed yellowish stipe, previously reported for *H. floccipes* var. *luteipes* Ortega & Zea described from Spain [70]; the latter is otherwise microscopically identical and of unknown phylogenetic status. Two unpublished reports of *H. floccipes* exist from Greece (D. Sofronis and G. Konstandinidis, pers. comm.).

#### *3.3. Litter and Other Terrestrial Decomposers*

Apart ECM and saproxylic fungi, several other mushroom species were recorded under black alder trees, and therefore constitute a part of the fungal diversity of the *A. glutinosa* priority habitat in Andros. Needless to say, none of these species is specifically linked to alders; instead they are considered 'generalists' to be found in both deciduous and coniferous forests. Among them, the following four species are recorded for the first time in Greece. *Lepiota ochraceofulva* P.D. Orton (Figure 8h) is a rather rare (but widespread) species in Europe, reported from *Fagus* and other deciduous trees on humus-rich, loamy soil [71,72]; it forms highly toxic mushrooms containing amanitins. Our single collection consisted of few basidiomes growing on a thick layer of leaf-litter under *A. glutinosa.* They are characterized by pilei of up to 7 cm in diam., with orange-brown scales; lamellae forming a distinct collarium and reddish-orange in maturity; spores ellipsoid to oblong, dextrinoid, not metachromatic in cresyl-blue, measuring 5.4–7.5(8.1) × 3.5–4(4.5) μm; basidia (2)4-spored, clamped; cheilocystidia short

clavate to cylindrical, rarely papilate, often in chains; pileipellis, a hymeniderm, composed of more or less clavate elements up to 50 μm long.

*Lepista ovispora* (J.E. Lange) Gulden (Figure 8i) is an uncommon (albeit widespread) European species recorded only once during this study on leaf litter under alders. Typical diagnostic features are the densely caespitose habit, the relatively fleshy basidiomes, the brown hygrophanous pileus with pruinose surface [73]. In addition, our specimens possessed spores ovoid to broadly ellipsoid, finely punctate, 4.7–6.8 × 3.8–4.4 μm, clamped basidia and no cystidia. *Psathyrella hellebosensis* D. Deschuyteneer, A. Melzer (Figure 8j) was recently described from Belgium [74] and was later reported from riparian alder habitats in Italy [75]. The morphological features of our collection are in agreement with the morphology of Belgian and Italian basidiomes, but since it corresponds to a rarely reported species, a detailed description of our material is hereby provided: pileus up to 3 cm in diam., hygrophanous from dark reddish-brown to greyish-beige, with scanty remains of veil; lamellae subdistant with whitish edge; stipe 2–4 × 0.2–0.3 cm, not rooting; spores 7.3–8.7 × 4.5–5.7 μm, Q = 1.43–1.73, ovoid to angular in face-view and not or weakly phaseoliform in side-view, not opaque; lamellae edge sterile composed exclusively of sphaeropendunculate paracystidia (no pleurocystidioid paracystidia were observed); pleurocystidia 34–48 × 11–17 μm, utriform. The material was collected from wet soil by the alluvial stream banks. This species shows high phylogenetic affinity to *P. thujina* A. H. Sm. by using ITS sequences only; however, it is clearly separated when the tef-1α marker is added in the phylogenetic analysis, while it is also distinguished by its distinctly larger and prominently phaseoliform spores [75].

Previous studies on the mycodiversity of Andros island reported the occurrence of four *Entoloma* species, one of them was new to science, i.e., *E. alnicola* Noordel. & Polemis [22,25]. Our recent field work resulted in other interesting collections of *Entoloma* spp. for which the identity and phylogenetic relationships to closely allied taxa are still under investigation. However, by using morphology alone, the presence of a rare European species was confirmed, namely *E. uranochroum* Hauskn. & Noordel. (Figure 8k) recorded for the first time in an alder habitat. This beautiful dark blue-violet mushroom was so far reported from subalpine meadows on calcareous soil in Austria (type locality) and the French Alps. Moreover, its striking microscopical features, e.g., the large fusiform cheilocystidida with granular yellowish-brown content, place it in the distinct section *Ramphocystotae* (Largent) Noordel., together with only one other European representative, namely *E. rhynchocystidiatum* Noordel. & Liiv [76].

#### **4. Conclusions**

A long-term study of the diversity of macrofungi in alder stands of Andros resulted in an inventory consisting of 106 species of basidiomycetes, including 21 taxa recorded for the first time in Greece. The majority of findings corresponded to saprotrophs (#92, mainly wood-rotting fungi) and the rest were ECM species. Considering the limited size of the area under study in a small Aegean island, the outcome of this work in terms of the number of taxa and variability is indicative of the wealth of the *A. glutinosa* priority habitat. However, the black alder stands in Andros have suffered considerably from floods in the past (as a consequence of fires that destroyed vegetation in the surrounding mountains which acted as a physical barrier protecting from downhill water runoffs) and their regeneration is hindered due to grazing by feral goats. The importance of fungi in the conservation/restoration of such natural habitats was demonstrated in the past [77,78], and recent activities focus at improving the status of the degenerated alder stands by exploiting indigenous ECM fungi as inoculants to young alder seedlings prior to their transplantation on site. Moreover, new knowledge about mushroom diversity and the ecological role of this group of organisms seems to enhance considerably people's perception and awareness, and hence facilitates implementation of conservations actions which are currently under way in selected alder stands of Andros.

**Supplementary Materials:** The following is available online at http://www.mdpi.com/1424-2818/12/6/232/s1, Table S1: Details of the 10 sampling sites in Andros island from where basidiomes were collected: locality name, coordinates, altitude (m a.s.l.) and surface of the study area (m2).

**Author Contributions:** Conceptualization, E.P. and G.Z.; methodology, E.P., G.I.Z., V.D. and V.F.; validation, E.P., V.D. and V.F.; formal analysis, E.P., G.I.Z. and V.F.; investigation, E.P., G.I.Z., V.D. and V.F.; data curation, E.P., G.I.Z. and V.F.; writing—original draft preparation, E.P.; writing—final draft, G.I.Z.; review and editing—final draft, E.P., G.I.Z., V.D. and V.F.; supervision, G.I.Z.; project administration, G.I.Z.; and funding acquisition, G.I.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by the project titled "Conservation of priority species and habitats of Andros Island protected area integrating socioeconomic considerations" (European Commission – LIFE-Nature, LIFE16 NAT/GR/000606).

**Acknowledgments:** We would like to thank V. Goritsas for the preparation of the map figures included in this work, and S. Adamcik, B. Dima, M. Noordeloos and J. Nuytinck for helpful discussions on some of the findings of this study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Genetic Structure and Phylogeography of** *Tuber magnatum* **Populations**

**Beatrice Belfiori <sup>1</sup> , Valentina D'Angelo 2, Claudia Riccioni 1, Marco Leonardi <sup>2</sup> , Francesco Paolocci <sup>1</sup> , Giovanni Pacioni <sup>2</sup> and Andrea Rubini 1,\***


Received: 31 December 2019; Accepted: 22 January 2020; Published: 24 January 2020

**Abstract:** The ectomycorrhizal fungus *Tuber magnatum* produces the white truffle appreciated worldwide for its unique aroma. With respect to other *Tuber* spp. of economic interest, *T. magnatum* presents a narrower geographical range. This species has, in fact, long been considered endemic to Italy. However, over the last few decades several reports have documented the presence of white truffles in different Mediterranean countries and in particular in various areas of south-east Europe. In this study, samples from several Pannonian and Balkan countries such as Hungary, Serbia, Romania, Bulgaria and Greece have been collected and genotyped with microsatellite markers and the data merged with those available for Italian populations. Our objectives were to test whether Italian and south-east European populations are differentiated and to evaluate the genetic diversity of *T. magnatum* all over its distributional range. We show the genetic structure of *T. magnatum* populations with the differentiation of four main groups: northern Italy, central-northern Italy, southern Italy and the Balkan/Pannonian region. The present study allowed us to refine the evolutionary history of *T. magnatum* and track the possible post-glacial expansion route of this species. The assessment of *T. magnatum*'s genetic structure is not only of scientific relevance, but it is also important for the conservation and market traceability of this prestigious fungus.

**Keywords:** microsatellite; SSR; white truffle; genetic diversity

#### **1. Introduction**

Species of the genus *Tuber* (Ascomycota, Pezizales, Tuberaceae) establish symbiosis with the roots of several tree and shrub species by forming structures for nutrients exchange, known as ectomycorrhiza [1]. In virtue of this mutualistic relationship these fungi produce hypogeous fruiting bodies (ascomata) known as truffles, that produce their spores sequestered within the surrounding tissues [2]. *Tuber* spp. thus rely on mycophagists for spore dispersal. Truffles produced by several *Tuber* spp. hold distinctive aromatic properties, which make them appreciated and marketed worldwide as food delicacies. Among edible *Tuber* spp. the black truffles harvested in Europe (*T. melanosporum* Vittad. and *T. brumale* Vittad.)*,* the black summer truffle *T. aestivum* Vittad., the whitish truffle *T. borchii* Vittad. and the white truffle *T. magnatum* Pico are of particular relevance.

The evaluation of the intraspecific genetic diversity and population genetic structure of a species is crucial to understand its biology and ascertain its origin, history and evolution. By using molecular markers and performing a wide geographical sampling, a fine assessment of the population genetic structure of *T. melanosporum, T. brumale, T. indicum, T. aestivum* and *T. magnatum* has been performed [3–9]. These studies highlighted the presence of geographically structured populations and phylogeographic patterns in these species.

Regarding *T. magnatum*, its genetic variability was initially investigated over a low number of specimens collected in Italy [10,11] or Italy and Croatia [12], using RAPD (Random Amplification of Polymorphic DNA), ITS-RFLP (Restriction Fragment Length Polymorphism of the Internal Transcribed Spacer of rDNA) and SNPs (Single Nucleotide Polymorphism) either in the ITS region, in the ß-tubulin gene or in a SCAR (Sequence Characterized Amplified Region). All these studies showed a very limited intraspecific polymorphism [13]. A broader investigation was carried out by Frizzi et al. [14] who analyzed the polymorphism of eleven isoenzymes on 139 specimens from 13 Italian populations. The low genetic variability across populations and the lack of an interpretable evolutionary trajectory were thought to be in agreement with a self-reproductive system, a restricted species endemism and a relatively recent differentiation of this taxon [14]. A few years later, by employing seven polymorphic simple sequence repeats (SSR) loci over 316 specimens from Italy and the Istrian peninsula (Croatia and Slovenia), Rubini and colleagues [4] disclosed for the first time an isolation by distance pattern and a phylogeographic structure in *T. magnatum,* with central Italy that likely represented a refugium for this species during the last ice age. In addition, this study has been instrumental for a deep reevaluation of the life cycle and the reproductive strategies of all *Tuber* spp. and to prove that truffle ascocarps are mainly made of the haploid, maternal tissue [15].

Around the late 90s of the last century, it became clear that *T. magnatum* can also be found in the Balkan peninsula and countries nearby [16] and, although more sporadically, in the south-east of France [17] and Switzerland [18]. The recent discovery of natural *T. magnatum* populations in areas ranging from Greece until Hungary and Romania ([13], and references therein), calls now for studies based on a larger sampling area than before. On these premises, here we employed SSR markers and an extensive sampling on most of the *T. magnatum* distributional range to shed more light on the genetic structure and phylogeography of this species. In particular, we aimed at evaluating whether Italian and south-east European populations are genetically differentiated and tracking the post-glacial expansion pattern of this species. The Balkan peninsula, like the Italian one, in fact could have represented a glacial refugia for *T. magnatum* during the last glaciation.

The assessment of the genetic diversity distribution could reveal important findings for aspects spanning from ecology, conservation and marketing of this prestigious fungus.

#### **2. Materials and Methods**

#### *2.1. Sample Source and DNA Analysis*

*Tuber magnatum* samples were collected in 2015–2016 with the help of local pickers. Sampling locations were mainly in Pannonian and Balkan countries and more marginally in central and southern Italy (Table 1). Genomic DNA was isolated from freeze-dried and fresh ascocarps (1–5 mg) according to Paolocci et al. [19]. DNA quantity and quality were evaluated using the spectrophotometer Nanodrop (MySpec, Wilmington, DE, USA). The isolated DNA was diluted to 20 ng/μL and stored at −20 ◦C. All samples were genotyped using eight microsatellite loci previously characterized in this species: the loci MA4, MA7, MA14, MA12, MA19 and MA13 derived from Rubini et al. [20] and the loci MA2-1 and MA5-1 derived from Rubini et al. [4]. The SSR loci were PCR-amplified using multiplex PCR [19]. To this purpose, two panels, each consisting of four loci, were defined (Table S1). PCR conditions were those reported in Rubini et al. [20]. Moreover, the locus MA13 was analyzed in all samples considered by Rubini et al. [4].

The SSR amplicons were analyzed by capillary electrophoresis using an ABI 3130 Genetic Analyzer in presence of the Genescan 500 LIZ size standard (Applied Biosystems, Foster City, CA, USA). Sizing of amplicons and allele scoring were performed using GeneMapper software version 3.7 (Applied Biosystems, Foster City, CA, USA).



*Diversity* **2020**, *12*, 44

\*

presence of missing data.

#### *2.2. Genetic Diversity Data and Population Structure Analyses*

The mean number of alleles (Na) and expected heterozygosity (He) for each locus and for each population were evaluated using GenAlEx v. 6.501 [21]. Allelic richness (Ar) was calculated with the software ADZE [22] using the rarefaction method [23] to correct for differences in sample size. The Ar was weighted to four individuals by excluding calculation for Population 1 and 31, which have a very small sample size. To avoid losing most of the information due to the small sample size of some populations, Ar was also calculated by sorting individuals into eight geographical groups according to their proximity (Figure S1). In this case, it was possible to consider a larger sample size, up to *n* = 15.

To estimate the degree of differentiation among populations, the analysis of molecular variance (AMOVA) and calculation of Fst and Rst values were performed using Arlequin software, version 3.5.1.2 [24]. These analyses were carried out both by comparing all 36 populations (sampling locations) and two regional groups of populations: Italy and Balkans/Pannonia.

Multilocus version 1.3b [25] was used to calculate the number of multilocus genotypes (MLGs) and genotypic diversity (i.e., the probability that two individuals taken at random have different genotypes). To evaluate if samples sharing the same MLG were true clones or resulted from random mating, the Psex (the probability of obtaining the same MLG from different sexual events) values and their significance levels were calculated with MLGsim software [26]. To test for the presence of an isolation by distance pattern, correlation between genetic and geographic distances was evaluated by performing a Mantel test according to Rousset [27] and using the software GenAlEx. The geographic distance matrix, consisting of the natural logarithm of the pairwise distance among populations, was calculated with GenAlEx considering the average geographic coordinates of each population. A genetic distance matrix consisting of pairwise Rst/(1−Rst) values was calculated using the software SPAGeDi version 1.5 [28].

The genetic structure of *T. magnatum* populations was evaluated by Bayesian analysis using the software STRUCTURE version 2.3.4 [29] and TESS version 2.3 [30,31]. Five independent runs of STRUCTURE for K (max number of estimated clusters) ranging from 2 to 10 were conducted. For each K, 2000,000 MCMC (Markov Chain Monte Carlo) and a burn-in of 200,000 iterations were performed, respectively. The admixture and no admixture models were tested considering correlated and uncorrelated allele frequencies. The optimal K was determined by comparing both the log-likelihood values and Evanno's ΔK [32] using the software Structure Harvester [33]. TESS analysis was performed both under the no admixture and the CAR admixture models by conducing 50 runs for each *K* ranging from 2 to 10 with 50,000 total MCMC steps and a burn-in of 10,000 sweeps. The spatial interaction parameter was set to the default value of 0.6 and the degree of trend to linear. To estimate the best K, the Deviance Information Criterion (DIC) was averaged across runs for each *K*. The smallest value before reaching a plateau was selected as the best K. Burn-in length and number of MCMC interactions for STRUCTURE and TESS were established by checking the convergence of summary statistics, and by evaluating consistence among runs of different lengths, following the recommendations in the software manuals.

STRUCTURE and TESS results were processed with the software CLUMPP [34] using the Greedy algorithm, random input order of runs and 1000 repeats. CLUMPP results were used to generate bar graphs using DISTRUCT [35]. The ancestry coefficients calculated with STRUCTURE were also plotted into a geographic map using the "POPSutilities" R script [36] according to the interpolation procedure described by Francois [37]. Interpolate values of ancestry coefficients among each pair of samples were calculated with R using the CLUMPP Q-matrix and a spatial grid obtained from the raster map of the area. For each sample, transition between one group (K) to another was displayed with a progressive change in the color intensity. The starting colors are those assigned to the different K.

#### **3. Results**

#### *3.1. Genetic Diversity of T. magnatum Populations*

SSR data were generated for 111 *T. magnatum* samples. Most of the samples were from the Balkan/Pannonian region (9 populations, 88 samples) and a few from central (Abruzzo-Molise, 18 samples) and southern Italy (Basilicata and Calabria, 5 samples) (Table S2). SSR analysis always showed the presence of a single allele per locus as expected for haploid organisms. In total, 49 alleles were detected and among them and 12 were new with respect to the alleles previously identified [4] (Table S2). The data obtained in this study were then merged with the data from Rubini et al. [4], which were relative to samples mainly collected in Italy, to end up with a dataset of 429 samples grouped into 36 populations covering almost all of the known *T. magnatum* distributional area (Table 1 and Figure 1).

**Figure 1.** Map showing the geographical location of the *Tuber magnatum* samples analyzed. Populations are indicated with circles and numbered as in Table 1.

Considering this merged dataset, the number of alleles was 77 with a minimum of 3 alleles for the locus MA13 and a maximum of 19 for the locus MA4. The expected heterozygosity (He) ranged from 0.12 (MA13) to 0.84 (MA51) (Table S3). In each population the mean number of alleles ranged from 1.4 to 4.6 and the expected heterozygosity from 0.17 to 0.56 (Table 1). Comparison of allele distribution between samples from the Italian and Balkan/Pannonian regions revealed 10 private alleles specific to Italy and 28 to the Balkans/Pannonia. Some of these 38 alleles were also private for single populations, but their frequency was no higher than 0.041 (Table S4). The remaining 39 alleles were shared between the two regions, the most of them without relevant differences. Only a few showed a biased frequency, this was the case of the alleles 162 and 172 at MA4 and MA14 loci, respectively, being markedly more frequent among the Balkan/Pannonian samples, and the allele 121 at locus MA21 more frequent among Italian samples (Table S4). By combining the allelic profiles at the eight SSR loci, 362 MLGs were identified. Among these MLGs, 48 were shared at least between two individuals, but only a few (9) turned out to be clones as they had Psex values that were significantly lower, thus rejecting the hypothesis of their origin by sexual reproduction (Table S5). These putative clones were detected only within a population. Some MLGs were also shared among both close and distant populations but none of them showed significant Psex values, suggesting that they were generated by chance because

of random mating (Table S5). The overall genotypic diversity was 0.998. When plotted against the number of loci, the genotypic diversity reached a value higher than 0.99 at six loci and also slightly increased up to eight loci (Figure S2).

The Ar for each single population ranged from 1.66 to 2.27. The highest values were observed for Abruzzo-Molise (Populations 4, 5 and 7) among the Italian populations and for Greece (Populations 35 and 36) among the Balkan/Pannonian populations (Table 1). Grouping individuals at a large geographical scale (i.e., 8 groups) allowed us to calculate the Ar by means of a rarefaction analysis based on a larger sample size (i.e., *n* = 15) and to show that the individuals from central-south Italy (Group 2) and those from the southern Balkans (Group 8) had the highest Ar values: 3.50 and 3.66, respectively (Figure S1).

#### *3.2. The Balkan*/*Pannonian and Italian Populations Were Genetically Di*ff*erentiated*

The presence of a genetic structure was revealed by AMOVA, which showed a marked and significant genetic differentiation between the 36 populations (Fst 0.169, *p* < 0.001; Rst = 0.422, *p* < 0.001). A higher significant differentiation (Fst 0.212, *p* < 0.001; Rst = 0.427, *p* < 0.001) was detected when two regional groups of populations (Italian and Balkan/Pannonian) were considered (Table S6). Furthermore, the Mantel test showed that the among-populations differentiation significantly increased with geographic distance (R<sup>2</sup> = 0.0232, *p* < 0.005; Figure 2).

**Figure 2.** Mantel correlation between the genetic and geographic distances.

To better evaluate the genetic structure of *T. magnatum* populations, an admixture analysis was performed. Bayesian clustering using the software STRUCTURE, showed that the most probable number of clusters (K) was four (Figure S3A,B).

Three of these clusters were primarily associated with different geographical areas of sample acquisition, with one grouping most of the individuals from southern Italy, one those from central-northern Italy and Istria, and the third those from Balkans/Pannonia. Conversely, neither a specific nor a prevalent geographical provenance emerged within individuals of the fourth cluster (Figure 3A). We also noted that some individuals of Balkans/Pannonia shared common ancestry with those of central-northern Italy, Istria and with Population 1 from southern Italy. Plotting the ancestry coefficients obtained with STRUCTURE into the geographical map produced a better picture

of the geographical distribution of individuals belonging to the four genetic clusters since they clearly matched the four distinct geographical areas: south Italy, central Italy, central-northern Italy and Istria as well as the Balkans/Pannonia (Figure 3B). No further relevant differentiation among individuals emerged when a higher value of K was considered (Figure 3A). Similar results were obtained using the no-admixture model considering either correlated or uncorrelated allele frequencies (data not shown). When STRUCTURE was run to analyze the Balkan/Pannonian regional group only, a further sub-structuration, with the differentiation of Greek Populations 35 and 36 starting from K = 3, emerged (Figure S4). Conversely, the same analysis on samples from Italy and Istria did not show any further sub-structuration (data not shown).

**Figure 3.** Bayesian analysis of genetic structure. (**A**) Plot of the ancestry coefficients of each single individual obtained with STRUCTURE based on admixture model and correlated allele frequencies. Each K is represented by a different color. Populations are indicated below the figure and their geographic origin above. (**B**) Spatial interpolation of population structure inferred for *K* = 4. Black dots represent the samples locations. The four K groups are represented by green, yellow, red and blue color gradients.

The genetic structure of populations was further explored using TESS. Unlike STRUCTURE, this software estimates the number of genetic clusters (K) by taking into account geographical coordinates of individuals to detect discontinuities in allele frequencies. Running TESS, under the admixture model there were three clusters matching; basically, the three main clusters found by STUCTURE (data not shown). The no-admixture model was more informative since, according to the DIC values, the most probable number of clusters was five (Figure 4A). In agreement with STRUCTURE, TESS evidenced a clear differentiation of southern Italian and Balkan/Pannonian populations. Moreover, the same analysis showed a further differentiation between populations of southern-central Italy and those of central and northern Italy. A fifth group was represented by the Balkan Population 35 from Greece and Population 1 from Calabria, which showed partial common ancestry (Figure 4B).

**Figure 4.** TESS analysis considering a no-admixture model. (**A**) Values of DIC in relation to the values of K. (**B**) Bar plots. Each individual is represented by vertical columns and the proportions of each K are represented by different colors. Populations and their geographic origin are given below and above the figure, respectively.

#### **4. Discussion**

When *T. magnatum* is evoked, what comes to mind varies according to people: According to the most gourmets and consumers, *T. magnatum* is the premium truffle, whereas for mycologists it represents the *Tuber* species that, among those of economic relevance, has the narrowest distributional range and the least understood autecology [13]. Previous studies revealed the presence of a genetic structure of *T. magnatum* Italian populations with southernmost and north-westernmost populations genetically differentiated from those of central Italy and Istria [4]. In these studies, only few samples other than those coming from Italy were considered; but, differently from what has been believed in the past, the *T. magnatum* distributional range is not indeed confined to the Italian peninsula only. Rather, it adds to the trans-Adriatic species as an increasing number of works have recently reported on the recovery of this fungus in different Balkan countries [16]. Thus, here we took advantage of *T. magnatum* specimens harvested from these new hotspots of white truffle production, spanning from Hungary, the northernmost, to Bulgaria, the easternmost, and down to Greece, to assess the genetic variability of this species and to ascertain whether genetic differentiation exists between populations of the two main regions of the *T. magnatum* distributional range: Italy and the Balkans/Pannonia. The analysis of more than 400 individuals, distributed into 36 populations located in Italy and the Balkan/Pannonian region and genotyped with eight SSR loci, unveils the presence of a high genetic variability and a clear genetic structure of *T. magnatum* populations.

#### *4.1. New Insights into Genetic Diversity of White Tru*ffl*e*

The study by Rubini and colleagues [4] conducted over more than 300 *T. magnatum* samples harvested in Italy and Istria and genotyped by means of a few SSR markers was the first to show the presence of genetically structured populations and the occurrence of an extensive gene flow within and among *T. magnatum* populations, likely thanks to spore dispersal by mycophagist mammals. With respect to previous studies, not only the higher sample size but also the use of the more informative markers, that is, polymorphic SSRs vs. isoenzymes, RAPD or SNPs on highly conserved loci (i.e., β tubulin), were therefore crucial to these authors to gain first evidence on the abundance and distribution of the genetic variability in this truffle species. Yet, the use of a handful of polymorphic SSR loci coupled to a sampling of *T. melanosporum* all over the distributional range was sufficient to Riccioni et al. [5] to confute the thesis of a trifling genetic polymorphism in this species [38] and map its possible post glacial recolonization pattern. According to this observation, here we tested the hypothesis that even a relatively small number of polymorphic SSR loci would be sufficient to shed light into *T. magnatum* genetic diversity, if samples representative of the entire species distributional range are analyzed. Following this rationale, the first goal of the present study was to broaden the *T. magnatum* sampling areas. Thus, despite the widely known and hard to counteract secrecy of truffle hunters, an extensive sampling of *T. magnatum* fruitbodies from Bulgaria, Hungary, Romania, Serbia and Greece has been performed. Eighty-eight samples from these countries and a few samples from south Italy have been genotyped for eight SSR loci and these data merged with the SSR profiles retrieved from Rubini et al. [4]. By doing so, the number of alleles detected among the 429 specimens increased up to 77, with 12 new alleles being found, and with the number of alleles per locus (from 3 up to 19) in the same range found for other truffle species, such as *T. melanosporum* (2–18) and *T aestivum* (4–15) [39,40].

The value of the He over loci of 0.54 (Table S4) or 0.38, if the average value among populations is considered (Table 1), is much higher than that reported for this species (0.16) when sampling was confined to southern Italy only [41]. This high value of He suggests that at large geographical scale the level of genetic diversity in *T. magnatum* is comparable with that of other European species with a wider geographical distribution, such as *T. melanosporum* and *T. aestivum,* which showed He values of 0.41 and 0.50, respectively [7,39]. The high level of genetic diversity is also confirmed by the high number of MLGs: 332 over 429 fruit bodies analyzed and the overall genotypic diversity of 0.99, a value similar to that found in *T. melanosporum* and *T. aestivum* [7,39]. Moreover, the evidence that genotypic diversity reaches a plateau value of 0.99 with just six loci indicates that the number of loci used in this study is adequate to evaluate the genetic variability of *T. magnatum*.

If the entire sample set is split into two main regions, the Italian and the Balkan/Pannonian ones, then the presence of many private alleles specific to a single region emerges, although with a very low frequency. It is worth mentioning that a few of the shared alleles show a quite different frequency between the two regions. In sum, enlarging the sample size with individuals from the easternmost and southernmost *T. magnatum* distributional areas has allowed us to disclose new and private alleles and gain a closer look into the population genetics of this species.

#### *4.2. Large Scale Sampling Reveals a Phylogeographic Structure in T. magnatum*

The AMOVA analysis of the SSR data obtained from the 429 samples proves the presence of a genetic structure among populations. The values of Fst and Rst, in fact, are higher in the present research vs. the previous work by Rubini et al. [4], suggesting that the new populations considered, mainly from the Balkans/Pannonia, are genetically differentiated from those of Italy and Istria. This is confirmed by the Fst and Rst values that not only are significant but also increase when two regional groups of populations (Balkan/Pannonia and Italy) are compared. In keeping with this, the Mantel test depicts an isolation by distance pattern (Figure 2).

The presence of a genetic structure has been evaluated more in detail by performing Bayesian analysis. The STRUCTURE algorithm clearly reveals that Balkan/Pannonian and Italian populations are genetically differentiated. Moreover, Italian populations are split into tree genetic clusters: southern, central and central-northern Italy, in agreement with results of Rubini et al. [4].

Historical population expansions and restrictions resulting from climate changes have been reported for plant species, including truffle host species, which experienced population bottlenecks as a consequence of glaciations [42]. In concert with this, the geographical distribution of European truffle species has followed the population expansion and restriction processes of their hosts [38,43]. For example, within the black truffle clade, *T. melanosporum* survived in refugia located in the Iberian and Italian peninsulas, as inferred by ITS, ISSR and SSR markers [3,5,44]. Conversely the surviving pattern of *T. brumale* aggr., resulting from the phylogeny of ITS, LSU and PKC loci, was more complex: Within the *T. brumale* clade A, populations of haplotype I survived the last glaciation in Western Europe, those of haploytpe II in Eastern Europe whereas those of clade B in the Carpathian basin and Balkan region [6]. This latter clade was later proposed as representing a cryptic species, *T. cryptobrumale* [45]. Concerning the *T. aestivum* clade, according to the distribution and relatedness of the ITS haplotypes of samples all over Europe and Turkey, it has been suggested that this species survived in Turkey whereas European populations likely experienced a population bottleneck during the last glaciation [9]. The geographic structure of *T. magnatum* populations from the Italian peninsula, as per SSR analyses, was consistent with the occurrence a glacial refugium in central Italy from which the northernmost and southernmost populations originated [4]. Thanks to the large sampling performed, this hypothesis is here reinforced as among Italian populations, those from the central-southern area exhibit the highest levels of allelic richness. Our study also suggests that the Balkan peninsula may have represented a *T. magnatum* glacial refugium as well. Our inference stems from the following considerations: (i) Balkan/Pannonian populations, with a few exceptions, belong to a different genetic cluster with respect to Italian specimens, suggesting an independent evolutionary history; (ii) STRUCTURE analysis performed in Balkan/Pannonian populations shows that the individuals from the southernmost populations (Greece) tend to differentiate from the others and the He allelic richness is higher in these populations, a situation frequently expected in correspondence to putative glacial refugia [46]; and (iii) many truffle host plant species, including those that host *T. magnatum* (i.e., beech and hornbeam spp.) survived the last glaciation in three main Mediterranean peninsulas: the Iberian, Italian and Balkan ones [47–49].

It is noteworthy that, although a general phylogeographic pattern emerged from the STRUCTURE analysis, some individual from the Balkan/Pannonian region (e.g., individuals from Populations 30, 32 and 34) share common ancestry with individual from Italian populations. Moreover, analyses that take in account the geographical information (i.e., TESS algorithm), confirm clustering into four groups but also show that the individuals from Population 1 (southern Italy) share partial ancestry with those from Population 35, one of the southernmost populations from the Balkan area.

The finding that individuals across the two shores of the Adriatic sea partially share a common ancestry poses the question whether strain migration occurred from one side of the Adriatic shore to the other. In fact, the hypothesis that, in the past, strains from southern-central Italy moved to the Balkan region or vice versa, cannot be ruled out. Many Mediterranean taxa present disjunct distributions between the west and east Mediterranean, and these disjunct biogeographical patterns are the results of the complex paleogeographic history of the present Mediterranean region [50]. Land bridges between the Italian and Balkan shores of the Adriatic sea occurred in the Neogene through the formation of the Apulo-Dalmatic Realm [51] and likely during the Pleistocene glaciations [52]. Thus, it is conceivable that by enabling the migration of mycophagist animals these geological events may have favored truffle spore dispersal between the two peninsulas. Along the same reasoning, as the two shores of the Adriatic sea shared basically the same climate and soil (calcareous soil of cretaceous origin) conditions [53], it is more than conceivable that whatever the truffle migration direction was, the newly introduced truffle strains encountered host species and pedoclimatic conditions that have favored their settlement. Mycophagist-mediated spore dispersal across the Alps also appears to be the most conceivable explanation of the higher relatedness of Istrian specimens to those from Italy rather than to those from the Balkans. A more extensive sampling of specimens from both sides of the Adriatic sea covering, in particular, the latitudes spanning from 40◦ to 42◦ N, coupled with the use of phylogenetically informative functional markers, would help us to further test the "bridge" hypothesis.

#### *4.3. Genetic Structure and Conservation Implications*

Unearthing *T. magnatum* population genetic structure may have important implications for conservation of its biodiversity. Differently from black truffle species, *T. magnatum* cultivation is not yet established [13]; thus, propagation of this species is almost exclusively natural. In vitro isolation of mycelium strains is also very challenging as this species shows a very slow growth rate and optimal nutritional requirements have not been identified yet. Thus, the preservation of *T. magnatum* strains ex situ in genetic banks is currently unfeasible. Rather, the most affordable strategy for the conservation of *T. magnatum* biodiversity relies on the preservation of its natural habitats. On these premises, results of the present study may be of relevance to identify and preserve populations and strains specific and adapted to different environments.

#### **5. Conclusions**

Here we have shown that *T. magnatum* genetic diversity is higher than hitherto thought and geographically structured across the Italian peninsula and Balkan/Pannonia region. Our findings are of relevance to make inferences about the phylogeographic history of this species but also for marketing and conservation purposes. The price of white truffles is traditionally dictated by their geographic provenance; thus, by increasing the number of genetic markers it would be possible, in the near future, to trace the origin of these truffles. This is a prerequisite to promote and sustain local white truffle-linked ecosystem services and economies but also to evaluate if and to what extent genetic determinants concur to shape the aroma variability across white truffles of different provenance. From a conservation point-of-view, the presence of a phylogeographic structure led us to hypothesize that *T. magnatum* strains of different geographic areas might exhibit different adaptation traits. In the light of the difficulties in its cultivation/propagation and of a global warming scenario, preservation of *T. magnatum* natural habitats from both Italy and the Balkan/Pannonian countries is therefore crucial to prevent the erosion of its biodiversity.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1424-2818/12/2/44/s1, Table S1: Panels of SSR loci used, Table S2: List of the samples considered in this study, Table S3: Polymorphism levels of the 8 SSRs over the entire sample set, Table S4: Allele distribution and frequency in samples from Italian (I) and Balkan/Pannonian regions (B), Table S5: List of MLG and results of MLGsim analysis. Populations are indicated when identical MLG are found, Table S6: AMOVA analysis among all populations and considering two regional groups of populations (Italy, and Balkans/Pannonia), Figure S1: Allelic richness in eight geographical groups. The geographical groups are indicated below the figure. Each geographical group includes all individuals from populations (numbered as in Table 1) reported in parentheses, Figure S2: Average genotypic diversity in function of the number of loci, Figure S3: Estimation of the most probable k. (a) Mean log likelihood over 5 runs (error bars = standard deviations) and (b) ΔK, the second order rate of change in the likelihood at each K, Figure S4: STRUCTURE analysis performed on Balcan/Pannonian populations only, based on admixture model and correlated allele frequencies. Each K is represented by a different color. Populations are indicated below the figure and their geographic origin above.

**Author Contributions:** Conceptualization, A.R., G.P. and F.P.; methodology, A.R., B.B. and C.R.; investigation, data curation, B.B., V.D., M.L. and C.R.; formal analysis, A.R. and B.B.; resources, G.P. and M.L.; writing—original draft preparation, A.R., G.P., F.P.; writing—review and editing, B.B., V.D., C.R., M.L., G.P., F.P. and A.R. All authors have read and agree to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** We are grateful to Pavlina Kladopoulou, Giorgio Konstanidis, Kiro Prodan, Kenan Kyose, Iordan Taralanski, Dejan Polic, Oszkar Fekete, Istvan Bagi, Silvio Guardiani, Gianni Miglietta, Mario Marchione, Marilena Oddis, Simona Ascione, Domenico Puntillo for providing us with fresh specimens of *Tuber magnatum* from the Balkans and Hungary and some areas of the Italian peninsula.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
