Our sampling aimed at covering the territory of the Czech Republic, and to a lesser extent adjacent parts of Slovakia, using two sources of material. First, for cultivation of axenic cultures and subsequent molecular analysis, 66 fresh specimens of Morchella that originated from our own recent collections or were provided by collaborating mycologists in 2008–2018, were used. Fruiting bodies from each micro-locality (unless significantly different in appearance) were considered as a single specimen and a single fruiting body was usually used for cultivation and/or analysis. However, at nine localities two to twelve mulch morels ascomata were analyzed to assess species diversity within the mulch beds (Suppl. material 1, Table S1). Cultures were derived either from the spore prints or from the inner tissues of ascomata transferred into Petri dishes with a malt extract glucose agar medium (MEGA; 10 g/L malt extract, 5 g/L glucose, 15 g/L agar) supplemented with chloramphenicol (100 mg/L). Cultivation was carried out in the dark at 18–20 °C. Axenic cultures of the obtained strains maintained on a rye grain substrate are available as a part of the Collection of Edible & Medicinal Macromycetes (CEMM) maintained within the framework of The Czech National Programme on Conservation & Utilization of Microbial Genetic Resources Important for Agriculture (http://www.vurv.cz/cspp/mikroorganismy/Edible and Medicinal macromycetes.html) at the Crop Research Institute (https://www.vurv.cz). For DNA extraction, the mycelium or sclerotia of individual strains were sampled from jars with the rye grain spawn prepared by the inoculation of pre-soaked and sterilized rye grains with pieces of agar covered with morel mycelium. From 16 specimens (including five samples from Slovakia) in which the derivation of axenic cultures was unsuccessful (marked as n.m.d. in Suppl. material 2, Table S2) DNA was extracted directly from fresh-frozen or dried pieces of ascomata.

Secondly, for the DNA analysis only, 377 morel specimens in total were obtained from the selected Czech public herbaria and one private fungarium, of which 203 were successfully analyzed (abbreviations according to Thiers 2018): BRNM: 73 specimens, CB: 50 specimens, CHOM: 23 specimens, HR: 24 specimens, LIT: 12 specimens, PL: six specimens, PRC: three specimens, and Vavřinec Klener’s private fungarium: 12 specimens). The specimens were collected between the years 1950 and 2018. For details see Suppl. material 2, Table S2.

Total genomic DNA was extracted from ca < 10 mg of dry fruiting body or an equivalent amount of the fresh mycelium culture or sclerotia by the CTAB method (Doyle and Doyle 1987). The ITS locus was amplified and sequenced in all the studied accessions using the ITS1F (Gardes and Bruns 1993) and ITS4 primers (White et al. 1990) or, in the case of old specimens with fragmented DNA, either with ITS1F and ITS2, or ITS3 and ITS4 (White et al. 1990). Subsequently, at least two representative accessions per detected phylospecies (with respect to the detected variation in ITS) and approximately six accessions within species-rich complexes were selected for further sequencing. RPB1 was amplified and sequenced with the gRPB1-A and fRPB1-C primers (Matheny et al. 2002), RPB2 with the fRPB2-7cF (Liu et al. 1999) and RPB2-3053r primers (Reeb et al. 2004), EF-1α with the EF-526F and EF1567R primers (Rehner and Buckley 2005), and domains D1 and D2 of 28S rDNA (LSU) with the NL1 and NL4 primers (O’Donnell 1993). All the PCRs were performed in 20-μL reaction mixtures with Kapa polymerase (Kapa Biosystems, Massachusetts, USA) and a touchdown protocol with an annealing temperature of 61–56 °C in the first six cycles and 56 °C in the following 37 cycles. The PCR products were purified by precipitation with polyethylene glycol (10% PEG 6000 and 1.25 M NaCl in the precipitation mixture) and sequenced by the Sanger method at Macrogen Europe (The Netherlands).

Sequences were edited and aligned in Geneious 7.1.7. (Biomatters, New Zealand) using the MAFFT plugin and deposited in NCBI GenBank under the accession numbers MH982584–MH983000. Alleles of the ITS locus were distinguished on the basis of single nucleotide polymorphisms and compared to the publicly available sequences. Bayesian phylogeny inference in MrBayes 3.2.4 (Ronquist et al. 2012) with 10⁷ generations, sampling every 1000^th tree, in two independent runs, each with six chains, 50% burn-in and a temp parameter of 0.01 was used for the preliminary assignment to the published phylogenetic species. As a reference database, previously published sequence data from one to two accessions per phylospecies were selected in order to cover the total species richness and the widest possible intraspecific variation. Multilocus sequences from the selected accessions were concatenated and ambiguously aligned parts (ITS1 in the Esculenta Clade dataset) and ends of the sequences with many missing data were discarded. Bayesian phylogeny inference for concatenated data was computed in MrBayes with 20 million generations, sampling every 1000^th tree, in two independent runs, each with four chains, and the first 10 million generations (50%) were excluded as burn-in. A substitution model for each locus was determined in Partitionfinder 2.1.1 (Lanfear et al. 2017) using the corrected AIC (AICc) and a greedy search, and partitions were subsequently set in MrBayes according to the loci.

The phylospecies concept of O’Donnell et al. (2011) was based explicitly on two main criteria. First, species were recognized if they were resolved as reciprocally monophyletic in at least one of the individual (i.e., single-locus) phylogenies and in the combined dataset (let us call this the “criterion of monophyly”), and second, if their genealogical exclusivity was not contradicted by analyses of any individual data partition (“genealogical criterion”). Understandably, none of these criteria can be fulfilled in species with a single collection. Therefore, the recognition of such species was based on a third criterion, i.e., genetic divergence from their sisters (“criterion of distinctness”). The three criteria are fully legitimate and intuitive. Nevertheless, whether they are met or not is always dependent on a particular dataset. After the addition of more collections (e.g., from different geographic regions), the criteria may cease to be fulfilled and the species thus become “illegitimate”. This seems to be the case of Mel-23 and Mel-24, which appeared to be reciprocally monophyletic and well differentiated originally, but several Czech specimens, as well as specimens previously determined as Mel-31 and Mel-32, share apomorphies with both of the lineages at each of the analyzed loci and, furthermore, exhibit several unique mutations. Their phylogenetic position therefore disrupts the clear distinctness of Mel-23 and Mel-24. At the same time, the Czech specimens do not form a separate lineage(s) and do not fulfill any of the three criteria (Fig. 1, Suppl. material 3, Fig. S1, Suppl. material 5, Table S3) and cannot be distinguished as separate phylospecies.

The criterion of monophyly and the genealogical criterion were not employed in many of the recent studies and were violated either consciously or because only one specimen was available. This is the case of, e.g., Mel-26 (M. deliciosa Fr.), Mel-31 (M. pulchella; see below), or Mel-33. Similarly, the criterion of distinctness seems to be considered only rarely and little attention seems to be paid to the genetic differentiation among newly distinguished phylospecies and the related ones in many studies. For instance, Mel-34 was distinguished as a separate lineage sister to Mel-20 (M. purpurascens) on the basis of a single specimen. Its genotype was similar or identical to Mel-20 at all the loci except RPB1, which provided an almost identical sequence to Mel-23/24/31/32 (M. pulchella; Suppl. material 3, Fig. S1, Suppl. material 5, Table S3, see also below). This discrepancy may represent a true biological signal, e.g. incomplete lineage sorting (Leliaert et al. 2014), but divergence between Mel-20 and Mel-23/24/31/32 is substantial (Fig. 1) and such an explanation is therefore questionable. This case illustrates how important it is to justify the distinction of new phylospecies by (i) the number of genetic differences at each locus and (ii) the inclusion of more than one sample in the analyses, which enables the effective exposure of base miscalling, erroneous alignment, PCR mutations, contaminations and other technical and processing errors that pose the high risk of introducing artifacts as new species (Thines et al. 2018). Beside the above-mentioned criteria, we therefore suggest that every newly distinguished phylospecies should be based on several (optimally three or more) different specimens (“criterion of minimal sampling”), and that it should differ from closely related species at most of the highly variable loci (in the case of morels, i.e., EF-1α, ITS, RPB1, RPB2) by at least one, but preferably more SNP’s that would be shared by all the individuals that are studied (“criterion of polygenic differentiation”). Although the latter criterion may be rather pragmatic and not fully reflect theoretical evolutionary processes at different loci, our analysis of average genetic differentiation shows that the closely related phylospecies differ at every locus by > 2 (but usually > 10) SNP’s (Suppl. matirial 5, Table S3). The criterion is therefore supported by empirical evidence and could also be useful for the potential recognition of new species in the future.

It is important to note that the proposed criteria should not be viewed as definite and insurmountable limits for taxonomy, but rather as a recommendation for cautiousness in introduction of new (phylo)species. Incomplete lineage sorting, hybridization, evolutionary stasis, and other factors may affect phylogenetic signal at each locus, particularly in recently diverged lineages (Mailund et al. 2014). Such closely related lineages may have already achieved reproductive isolation and segregated in distinct ecological or biogeographical compartments, they may even have acquired some diagnostic morphological traits, but DNA phylogenies may fail to assign them to (reciprocally) monophyletic clades (reviewed by Leliaert et al. 2014). Final taxonomic decision may therefore be influenced by stronger lines of evidence than (weak) patterns in DNA sequence variation. Nevertheless, as stressed by Carstens et al. (2013: 4369), taxonomic inferences should be conservative, “for in most contexts it is better to fail to delimit species than it is to falsely delimit entities that do not represent actual evolutionary lineages”.

According to our analyses, the phylogenetic lineages of morels occurring in the Czech Republic are as follows (arranged by the phylospecies designations):

Mel-3 (M. semilibera DC.; Fig. 5A). Seventeen Morchella cultures or exsiccated specimens (one originating from Slovakia) were proven to be Mel-3, corresponding to Morchella semilibera (syn. Mitrophora semilibera (DC.) Lév., Morchella gigas (Batsch) Pers. or M. hybrida Pers.), which is in accordance with the previous phenotypic determination of the specimens. Ascomata were collected from mid-April to mid-May. As morphological features seemed to be highly reliable for the delimitation of this species (there is only the possibility of confusion with Verpa bohemica (Krombh.) J. Schröt.), fungarium specimens were mostly not used for DNA analyses. However, collection data for 50 M. semilibera fungarium specimens were included on a map (Fig. 3) to demonstrate the species distribution in the Czech Republic. Morchella semilibera is a widely distributed Eurasian species that had previously been recorded not only from the Czech Republic, but its occurrence was also confirmed molecularly from France, Germany, Italy, the Netherlands, Spain, Sweden, Turkey, and India (Taşkın et al. 2010, 2012; Kanwal et al. 2011; O’Donnell et al. 2011; Clowez 2012; Du et al. 2012b; Richard et al. 2015).

The Czech collections originated mostly from (semi-)natural habitats such as deciduous or, less frequently, mixed forests and floodplain forests (note that most of what are termed forests in Central Europe are semi-natural or completely artificial), groves, old fruit orchards, shrubs, or rocks. Only rarely was M. semilibera found in urban areas, e.g., in gardens, town parks, or also in ornamental beds, but we have no information as to whether there was bark mulch or not. The species appeared most frequently in association with Fraxinus spp., Carpinus betulus, Quercus sp., Acer spp., Prunus spp. (especially P. spinosa), and cherry trees. According to the literature, M. semilibera often grows under Fraxinus excelsior (Clowez 2012; Richard et al. 2015), and it was also found under Malus sylvestris, Castanea sp., and Populus sp. (Taşkın et al. 2010, 2012; Richard et al. 2015). Judging by our recent collections and the representation of the species in herbaria, M. semilibera seems to be one of the most common Morchella species and is widespread in lowland areas of the Czech Republic (Fig. 3). However, in the national red list of macromycetes it is treated as Near-Threatened because of the potential overexploitation of natural populations by mushroom gatherers (Antonín 2006).

Figure 5. 

Examples of the fruiting bodies of Morchella phylospecies in the Czech Republic. A1–2 Mel-3 (M. semilibera; A1. accession number VK13, A2. IP229) B Mel-10 (M. importuna; IP26) C Mel-13/26 (HR86151) D Mel-15/16 (IP245 and IP247) E Mel-19 (M. eohespera; HR99241) F Mel-20/34 (HR102132) G Mel-23/24/31/32 (HR102133) H Mel-39 (VK17) I Mes-4 (M. americana; IP297) J Mes-5 (IP350, IP351) K Mes-8 (M. esculenta; IP341) Photographers: Vavřinec Klener (A1, H, K); Irena Petrželová (A2, B, D, I, J); Jan Kramoliš (F, G); Dušan Bureš (C, E).

Mel-10 (M. importuna M. Kuo, O‘Donnell & T.J. Volk; Fig. 5B). Mel-10 is a newly recognized species for the Czech Republic, although our results show that it has already been a part of the Czech mycobiota for decades. In total, 70 Morchella cultures or exsiccated specimens (the oldest one was collected in 1950) previously morphologically identified mostly as M. pragensis (18 specimens), M. conica (12 specimens), or M. elata (nine specimens), or designated just as “black mulch morel” (18 specimens) were determined as Mel-10. The ascomata of the Czech collections were extremely variable in shape. The specimens that were examined were collected from mid-April to mid-May. It is noteworthy that M. importuna is probably a later synonym for several validly published names, e.g., M. elata, M. hortensis Boud. or M. vaporaria Brond. However, the interpretation of these names is unresolved and the name M. importuna was therefore provisionally retained for Mel-10 by Richard et al. (2015). Among Czech mycologists and morel hunters, the species is often treated as M. pragensis. This name was published twice (Smotlacha 1947, 1952) on the basis of collections from the surroundings of Prague, firstly without the Latin diagnosis, secondly without the holotype being indicated (which, nevertheless, was not necessary before 1958) and as two forms without the nominate form being specified. The nomenclatural errors were later corrected and the neotype was assigned by Moravec (1970). However, sequencing of the neotype has not been successful yet; its identity needs to be determined in future studies. Despite the formal errors, M. pragensis became widely known among the public and the name has been commonly used for various collections from anthropogenic habitats, particularly ruderal places such as waste dumps and debris after demolition, but also gardens, yards, ornamental beds with bark mulch, etc.

Morchella importuna was described from the USA in 2012 (Kuo et al. 2012) and it was hypothesized as originating in western North America, from where it has spread in association with horticulture and silviculture (Taşkın et al. 2010), but it has also been reported from Germany, Poland, Finland, France, Switzerland, Spain, Turkey, Cyprus, Israel, Canada, and China (Taşkın et al. 2010, 2012; Du et al. 2012a, 2012b; Richard et al. 2015; Loizides et al. 2016). The species appears to be a saprotroph (Mann and Mann 2014), and therefore it can be cultivated artificially (Du et al. 2015). Morchella importuna is also known as a facultative post-fire species (Clowez 2012; Du et al. 2012a; Loizides 2017), and recent research has shown that it can even be grown on fire-treated fields (Li et al. 2017). A morphological description of M. importuna was given by Clowez (2012; under the name M. vaporaria) and Kuo et al. (2012). It may be difficult to distinguish M. importuna from other species in the Elata Clade morphologically. However, the best clue for its identification may be its occurrence in urban habitats in combination with regularly laddered, vertically oriented pits and ridges on ascomata (Kuo et al. 2012; Mann and Mann 2014). Both in its presumed native distribution area and in the Czech Republic it occurs in various urban habitats, particularly woodchip or mulch beds (Kuo et al. 2012); therefore, it is sometimes called the “mulch morel” (Mann and Mann 2014). But it is also frequently found in the yards of houses, in masonry, dumps of rubble, sand, wood or bark, and one Czech specimen was collected in an old fire pit. Only occasionally was this species found in semi-natural habitats such as forests, along forest paths, or in meadows or town parks.

Mel-13/26 (M. deliciosa Fr.; Fig. 5C). Eight Morchella cultures or fungarium specimens phenotypically identified mostly as M. conica (four specimens) clustered with both Mel-13 (no Latin binomial) and Mel-26 (M. deliciosa). While the former species was distinguished by O’Donnell et al. (2011), the latter one was delimited by Taşkın et al. (2010), but without Mel-13 being used in their analysis. It was only later that Taşkın et al. (2012) analyzed both species together and revealed the paraphyly of Mel-13 because of the exclusion of Mel-26 from the clade. The separation of Mel-26 can therefore be considered inappropriate and both species were combined by Du et al. (2015) as Mel-13/26 (M. deliciosa being the only known Latin binomial). Our results confirm this treatment unambiguously. Mel-13 has so far only been reported from Asia (China, India, and Turkey; O’Donnell et al. 2011; Du et al. 2012a; Taşkın et al. 2012; Richard et al. 2015), while Mel-26 has, in addition to Turkey (Taşkın et al. 2010, 2012), also been reported from some European countries (France, Poland, and Sweden; Clowez 2012, as several varieties of M. conica; Taşkın et al. 2012; Richard et al. 2015; Baran and Boroń 2017). The Czech specimens that were examined were collected in mid-April in mixed forests, mainly under Fraxinus excelsior, Picea sp., and Pinus sp. Other nearby trees or shrubs were Quercus sp., Larix decidua, Fagus sylvatica, and Sambucus nigra. Other authors (Taşkın et al. 2010, 2012; Clowez 2012) mainly reported associations with conifers (L. decidua, Picea abies, Pinus sp.), while Baran and Boroń (2017) also observed Abies alba, Tilia platyphyllos, Acer pseudoplatanus, and Euonymus verrucosa at the localities where M. deliciosa was collected.

Mel-15/16 (M. angusticeps Peck / M. eximioides Jacquet.; Fig. 5D). Nine Morchella cultures or fungarium specimens previously morphologically recognized mostly as M. conica (six specimens) clustered with Mel-15 (M. angusticeps) or Mel-16 (M. eximioides). Both species were originally delimited by O’Donnell et al. (2011) on the basis of a sample set of seven eastern North American (Mel-15) and four Scandinavian specimens (Mel-16), which exhibited stable polymorphism (i.e., synapomorphies) at EF-1α and RPB2, but not at LSU and RPB1. Nevertheless, the Czech samples share synapomorphies with both of the species and form a basal lineage to them (Fig. 1). The Czech Mel-15/16 specimens fulfill the criterion of monophyly and the genealogical criterion (Suppl. material 3, Fig. S1) and thus could be distinguished as a separate phylospecies according to O’Donnell et al. (2011). However, the three lineages are distinguishable only by EF-1α and RPB2 (Suppl. material 3, Fig. S1, Suppl. material 5, Table S3), and the total detected genetic distance of the Czech specimens from Mel-15 and Mel-16 is eight and 15 SNP’s, respectively. Moreover, variation within branches is higher than that among branches at some loci and the split into three lineages therefore may be caused, hypothetically, by geographic variation and limited sampling. Consequently, we prefer not to assign the Czech lineage as a new phylospecies before additional (e.g., phenotypic) data prove its distinctness.

It is generally supposed that Mel-15 is endemic to eastern North America (O’Donnell et al. 2011; Kuo et al. 2012; Richard et al. 2015), while Mel-16 has been reported from Northern Europe and China (O’Donnell et al. 2011; Du et al. 2012a). This early fruiting species (recent collections were made in approximately mid-April) appeared in both (semi-)natural and anthropogenic habitats. Five specimens were found in deciduous forests, a town park or an old orchard. Two other specimens were collected in the vicinity of paper mills on paper or wood waste. Two specimens (collected in different years) were found by a sedimentation basin of a heating plant. The species grew together with various deciduous trees and shrubs, including Prunus spp., Fraxinus sp., Acer sp., Populus sp., Crataegus sp., Cornus sanguinea, Betula sp., or Salix sp. Du et al. (2012a) also reported the association of Mel-16 with Picea sp.

Mel-19 (M. eohespera Beug, Voitk & O‘Donnell; Fig. 5E). Forty-three Morchella cultures or fungarium specimens morphologically identified mostly as M. conica (24 specimens), M. elata (seven specimens), or M. pragensis (six specimens) were determined as Mel-19 (M. eohespera; an older name is possibly M. norvegiensis Jacquet.; see Voitk et al. 2016b). Although Mel-19 was previously recorded from the Netherlands, Sweden, Switzerland, China, and the USA (O’Donnell et al. 2011; Du et al. 2012a; Taşkın et al. 2012; Beug and O’Donnell 2014; Richard et al. 2015), it was only after the collections from Canada that the Latin binominal was given to this phylogenetic species (Beug and O’Donnell 2014; Voitk et al. 2016a). A morphological description of this species is available in Voitk et al. (2016a). In the Czech Republic, M. eohespera appeared from mid-April to mid-May, but the most recent collections were mostly made around mid-April. The specimens that were examined were collected in a variety of habitats including different types of forests (where they often occurred along roads, on deposits of wood on wood waste, on hillsides, and in river or creek valleys), in gardens, old yards, rubble sites, railway stations, and other urban habitats, sandstone quarries, a brick factory, a meadow, and also on bark mulch. Voitk et al. (2016a) also reported the occurrence of this species both in natural habitats and at sites significantly affected by human activities. The Czech collections of M. eohespera were frequently found together with Populus spp. (mostly P. tremula), Betula pendula, Picea abies, or Pinus sylvestris, while other nearby trees were Salix spp., Fagus sylvatica, Fraxinus excelsior, Quercus sp., and Malus domestica. The species can also be found in association with other trees such as Alnus sp., Corylus sp., or Abies sp. (Du et al. 2012a ; Voitk et al. 2016a).

Mel-20/34 (M. purpurascens (Boud.) Jacquet; Fig. 5F). One culture and 21 fungarium specimens previously morphologically identified mostly as M. conica (11 specimens) or M. elata (six specimens) were determined as Mel-20, corresponding to M. purpurascens, or as Mel-34, which lacks a Latin binominal. Mel-20 was originally distinguished by O’Donnell et al. (2011) as a sister lineage to Mel-19 (M. eohespera) with a very low bootstrap support at most of the loci (< 50%) and 93% support at EF-1α, which contains almost all of the few apomorphies that distinguish the two lineages. Later, Du et al. (2012a) distinguished Mel-34 on the basis of a single specimen from China that is almost identical to Mel-20 at all loci except RPB1, which provided an almost identical sequence to Mel-23/24/31/32 (M. pulchella; Suppl. material 3, Fig. S1, Suppl. material 5, Table S3). Therefore, we propose merging Mel-34 with Mel-20 (M. purpurascens being the only known binomial) provisionally until more collections are made and this extraordinary pattern is confirmed. The distinctness of Mel-19 and Mel-20 is still clear after the inclusion of the Czech samples, although the difference is very small, the latter branch is poorly supported (posterior probability 0.74) and distinction of the two lineages may thus be an artifact of anagenesis in the EF-1α gene. Not surprisingly, distinguishing between these species on the sole basis of ITS is tricky, as they differ in a single SNP. Moreover, this polymorphism is not stable across the whole ranges of these species, but it was stable for all of the Czech specimens that were studied. We therefore considered it as diagnostic in ITS-based determinations.

Mel-20 (M. purpurascens) is known from France, Scandinavia, Turkey, China, and Taiwan (Taşkın et al. 2010, 2012; O’Donnell et al. 2011; Clowez 2012, as a variety of M. conica; Du et al. 2012a; Richard et al. 2015). The ascomata of the Czech specimens were collected between late April and mid-May, both in anthropogenic habitats (railways, roadsides, gardens, and also in a junkyard, places with deposits of various materials such as rubble, sand, wood, or pure brick clay, between stones or even on concrete) and in forests, and co-occurred with both conifers (often growing under Picea sp. or Pinus sp.) and deciduous trees and shrubs (Quercus sp., Prunus domestica, Betula sp., Crataegus sp., Populus sp., Salix sp.). Other authors also reported the frequent co-occurrence of Mel-20 with conifers such as Abies sp., Pinus sp., or Cedrus sp. and also with Populus sp. or Quercus sp. (Taşkın et al. 2010, 2012 ; Clowez 2012 ; Du et al. 2012a).

Mel-23/24/31/32 (M. pulchella Clowez & F. Petit; Fig. 5G). Twenty-five specimens (including two samples from Slovakia) originally determined mostly as M. conica (10 specimens), M. elata (four specimens), or M. pragensis (five specimens), grouped together with Mel-23, Mel-24, Mel-31, and Mel-32. Mel-23 (no Latin binominal) and Mel-24 (M. septentrionalis M. Kuo, J.D. Moore & Zordani) were originally distinguished by O’Donnell et al. (2011) on the basis of three Scandinavian specimens and one specimen (plus five additional that were not shown) from the eastern USA and Canada, respectively. Mel-31 (M. pulchella) was delimited in parallel by Taşkın et al. (2010) without the inclusion of Mel-23 and Mel-24 in the analyses and has so far been reported from China, Pakistan, Turkey, and France (Taşkın et al. 2010, 2012; Du et al. 2012a; Richard et al. 2015; Badshah et al. 2018). Mel-32 (M. conifericola Taşkın, Büyükalaca & H.H. Doğan) was later distinguished by Taşkın et al. (2012) in Turkey, although the clade was poorly supported (BS = 59 for concatenated data and < 50% for individual loci). Moreover, the latter study revealed the paraphyly of Mel-31 as a result of the exclusion of Mel-24, and also very low support (if any) for each of the four species. These facts were confirmed by Richard et al. (2015) and our own data (Fig. 1). Therefore, we suggest combining the four formerly delimited species and to treat them as one, with the oldest known Latin binomial being M. pulchella, and M. septentrionalis and M. conifericola being later synonyms. Two specimens corresponding to the Mel-23/24/31/32 lineage were also reported from India (Du et al. 2012b). The degree of endemism, therefore, appears to be overestimated and the recorded variation may be attributed to phylogenetically young mutations and sometimes to intraspecific geographic variability. The Czech specimens were collected from mid-April to early May, mostly in forests (often along roads or forest edges), but also along railways or in sandstone quarries. A special collection site was a surface coal mine where a stable population of morels was visited in several successive years in the 1970s. Ascomata were mostly found under deciduous trees such as Populus tremula, Carpinus betulus, Betula pendula, or Fraxinus sp., while other nearby trees were Tilia sp., Salix caprea, Quercus sp., and occasionally conifers (Pinus sp., Picea sp.).

Mel-39 (newly designated phylogenetic species; Fig. 5H). Two cultures and four exsiccated specimens formed a well-separated and well-supported lineage sister to Mel-10 (M. importuna). Although closely related to Mel-10, this lineage differs at each of the loci that were studied by three (LSU) to 18 (EF-1α) synapomorphic SNP’s (59 SNP’s in total; Suppl. material 5, Table S3). This lineage is identical to the New-2 clade sensu Du et al. (2012b), which was reported from China and Germany based on ITS only; however, it has not been definitely distinguished as a phylospecies until now. Beside the significant genetic differences and genealogical concordance, Mel-10 and Mel-39 differ in several phenotypic traits, particularly on sclerotia under experimental cultivation. Whereas the sclerotia of Mel-39 are very tiny (mostly not bigger than 1 mm), spherical, not coalescing, dark, red-brown, mature sclerotia of M. importuna are of a light color varying from that of a walnut shell to somewhat orange, and coalesce into big hardened bodies of an irregular shape that are up to several centimeters long (Petrželová unpublished data). Nevertheless, formal taxonomic treatment needs to be based on extensive phenotypic analyses and the study of the type material of related taxa, and cannot be performed at this stage. The species was mostly collected on bark mulch, mostly in late April or early May, i.e., similarly to or slightly earlier than Mel-10.

Mes-4 (M. americana Clowez & Matherly; Fig. 5I). One Morchella strain maintained as an axenic culture and previously morphologically determined as M. esculenta was identified as Mes-4 (Fig. 2; Suppl. material 2, Table S2). This species was described under several binominals (Clowez 2012; Kuo et al. 2012), and M. americana was selected by Richard et al. (2015) as the most appropriate among the priority names. However, Clowez (2012) used the name M. rigida (Krombh.) Boud. for the French specimens that Richard et al. (2015) found as conspecific with Mes-4. If this name of Krombholz (a 19^th-century Prague mycologist) was used correctly, M. rigida (basionym M. conica var. rigida; Krombholz 1831–1834) is probably the oldest name for the species. Nevertheless, we follow the latest treatment of Richard et al. (2015) for now.

Morchella americana appears to be native to North America, where it is the most widely distributed Esculenta clade species (O’Donnell et al. 2011; Du et al. 2012b; Kuo et al. 2012; Richard et al. 2015). To date, it has also been reported from France, Spain, Turkey, and China (Taşkın et al. 2010, 2012; Du et al. 2012b; Richard et al. 2015), and this is the first record of the species for the Czech Republic (not considering Krombholz’s collections). In North America it mostly co-occurs with Fraxinus spp., Ulmus americana, Populus spp., Platanus occidentalis, Acer sp., or Quercus spp., but it can also be found in old apple orchards and occasionally together with conifers (Kuo et al. 2012; Richard et al. 2015). Its association with Buxus sempervirens has also been reported (Clowez 2012; Richard et al. 2015). It has been suggested that M. americana has only recently been introduced to Europe, as most records come from sites with a strong anthropogenic influence, especially from hybrid poplar plantations (Richard et al. 2015). The Czech specimen, nevertheless, occurred in the Žebračka National Nature Reserve, i.e., a site with one of the most valuable natural alluvial forests in the Czech Republic. Moreover, if Krombholz’s specimens originating from Czech Republic were really identical with Mes-4, human-mediated introduction from North America would be rather unlikely. The studied specimen was found under young Populus tremula trees at the beginning of May.

Mes-5 (Fig. 5J). Nineteen Morchella cultures or exsiccated specimens (including two samples from Slovakia) that had previously been phenotypically determined differently, mostly as M. vulgaris (eight specimens) or M. esculenta (three specimens), but also as a variety of species of black morels, were identified as Mes-5 (no Latin binominal). Although the multi-gene approach was only used for two specimens, no clear polymorphism was found at ITS among our Mes-5 accessions, whereas two SNP’s were observed between Mes-5 and the sister Mes-17 (M. dunensis). Therefore, identification based on ITS should be sufficient. Mes-5 has so far been found in Denmark, France, and Norway (O’Donnell et al. 2011). The Czech specimens were mostly collected from mid-April to early May, often in gardens (also on bark mulch) or in forests, but some collections were also made in a park, an orchard, a meadow and a waste dump near a summer cottage. Ascomata of this species were often found under fruit trees (Malus domestica, Pyrus sp., Prunus persica), and Rosaceae shrubs (e.g., Crataegus sp.); other nearby trees that were recorded were Fraxinus excelsior, Acer pseudoplatanus, Populus sp., Robinia pseudoaccacia, and Picea abies, and in Slovakia also Swida sp. and Pinus nigra.

Mes-8 (M. esculenta (L.) Pers.; Fig. 5K). Forty-seven Morchella cultures or exsiccated specimens, (including three samples from Slovakia) previously morphologically determined mostly as M. esculenta (29 specimens) or less frequently as M. crassipes (eight specimens) or M. vulgaris (six specimens), were identified as Mes-8 (corresponding to M. esculenta). Morchella esculenta is the common and widely distributed European morel species recorded from the Czech Republic, Poland, Germany, Switzerland, France, Spain, Belgium, the Netherlands, Norway, Sweden, and Turkey (Taşkın et al. 2010, 2012; O’Donnell et al. 2011; Clowez 2012; Du et al. 2012b; Richard et al. 2015; Baran and Boroń 2017) but also from China (Du et al. 2012a). The Czech collections were mostly obtained from mid-April to early May, occasionally up to mid-May, mostly in (semi-)natural habitats in deciduous (including floodplain) forests and shrubs, less often in limestone quarries, old orchards, parks, or gardens (here also on bark mulch). A special collection site was the edge of a reed bed. Uncommonly, M. esculenta was found in ruderal or urban habitats such as the yards of buildings, a rubble site or even the concrete floor of a woodshed. M. esculenta has been found in association with a variety of deciduous wood species. The Czech collections were more frequently collected under Fraxinus sp. (often F. excelsior), Crataegus sp., Prunus spp. (especially P. spinosa, P. domestica, and P. avium), Quercus spp., and Acer spp. Other nearby trees were Alnus sp., Carpinus betulus, Populus tremula, Betula sp., Salix caprea, Malus domestica, Aesculus hippocastanum, Robinia pseudoaccacia, and occasionally conifers such as Picea sp., Larix sp., Pinus sylvestris, or Thuja sp. According to other authors, M. esculenta can also grow under Ulmus minor, Malus sylvestris, Cydonia oblonga, Mahonia sp., or, rarely, Cupressaceae species (Clowez 2012) or Abies sp. (Taşkın et al. 2012).

What are known as the “mulch morels” represent a specific ecological group of morels that occur massively in newly created ornamental beds with bark mulch, mostly in gardens or around construction zones and newly built houses. On the basis of our observations, the macromorphological variation both within and among populations of mulch morels is remarkable, sometimes to such an extent that it brings to mind the variation among species. Therefore, we aimed at an estimation of the number of species within and among neighboring localities. However, with the only exception of one site with Mel-10 (M. importuna) and Mel-39, all the samples from the same ornamental bed belonged to the same species (Mel-10; Suppl. material 1, Table S1). Nevertheless, among the total of 48 specimens that originated from different localities with bark mulch, five Morchella species were recognized. A total of 36 specimens were determined as Mel-10 (M. importuna), five as Mel-39, four as Mes-5, two as Mel-19 (M. eohespera), and one as Mes-8 (M. esculenta; Suppl. material 2, Table S2).