Introduction
Species of Umbilicaria Hoffm. are predominantly saxicolous lichens, mostly found in regions of higher latitudes or altitudes worldwide (Frey Reference Frey1936a, Reference Freyb; Llano Reference Llano1950; Wei & Jiang Reference Wei and Jiang1993). Umbilicaria subg. Papillophora Davydov et al. is phylogenetically the most basal subgenus and includes species with a mostly Holarctic distribution, and European, East Asian and South American endemic elements (Davydov et al. Reference Davydov, Peršoh and Rambold2017, Reference Davydov, Ahti and Sennikov2020a, Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayerb). Two morphologically different species groups were segregated in the subgenus Papillophora, the U. vellea group and the U. angulata group. Species of the U. vellea group in this broad sense include well-known species, such as U. vellea (L.) Michx., U. cinereorufescens (Schaer.) Frey, U. crustulosa (Ach.) Lamy and U. spodochroa Hoffm.; they commonly possess rhizinomorphs that may or may not bear multicellular thalloconidia, with different types of apothecia having 8-spored asci with simple or submuriform ascospores.
Umbilicaria vellea was described by Linnaeus (Reference Linnaeus1753). Other species within the U. vellea group were described by lichenologists in the 18th and 19th centuries. Due to uneven knowledge at the time, many taxa were poorly understood and their species boundaries were debated (e.g. Hoffmann Reference Hoffmann1790, Reference Hoffmann1794; Acharius Reference Acharius1803; Merrill Reference Merrill1906; Hestmark Reference Hestmark2007; Davydov et al. Reference Davydov, Ahti and Sennikov2020a). Modern concepts for many species of the U. vellea group were substantiated by Frey (Reference Frey1929, Reference Frey1931, Reference Frey and Zahlbruckner1933, Reference Frey1936a, Reference Freyb) who provided an artificial key and detailed descriptions that are still valuable. In particular, Frey (Reference Frey1929, Reference Frey1931) recognized U. cinereorufescens as a distinct species. Poelt & Nash (Reference Poelt and Nash1993) studied U. vellea and morphologically similar species and emphasized that Umbilicaria are particularly difficult to interpret because multiple individual thalli covering a range of ages must be collected to understand the full developmental sequence. Based on differences in the shape of rhizinomorphs, Poelt & Nash (Reference Poelt and Nash1993) described U. americana Poelt & T. H. Nash from North America, so far included within the concept of U. vellea. East Asian U. orientalis Davydov is another species in the group that resembles U. vellea but differs in terms of rhizinomorphs and thalloconidia (Davydov et al. Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayer2020b).
Although Poelt & Nash (Reference Poelt and Nash1993) commented that there is no obvious reason to assume that more than one taxon exists within Umbilicaria vellea in Europe, we discuss morphological and molecular phylogenetic evidence to describe a new species, U. ahtii Davydov, from the U. vellea species complex. Umbilicaria ahtii has a fairly long history of understanding and informal acceptance. Back in 2003, during work on the Umbilicariaceae phylogeny (Davydov et al. Reference Davydov, Peršoh and Rambold2017), ITS sequences assigned to U. vellea were obtained by ED. Sequences formed two clearly distinct clades: most of them (‘U. vellea-1’) appeared as a sister of U. hirsuta (Sw. ex Westr.) Ach., while the others (‘U. vellea-2’) clustered as a sister group to U. cinereorufescens. The same result was obtained after re-extracting DNA, increasing phylogenetic markers and including additional species. In the final publication (Davydov et al. Reference Davydov, Peršoh and Rambold2017) based on ITS, mtLSU and RPB2, ‘U. vellea-1’ was found to cluster with U. hirsuta, and ‘U. vellea-2’ with U. meizospora (Harm.) Davydov & D. Masson, suggesting the presence of an undescribed species in Northern Europe. Examination of morphology revealed only minor differences in rhizomorphs and the colour of upper and lower surfaces, which are highly variable traits and not always discernible in scarce herbarium collections. To identify diagnostic features, extensive collections were studied in parallel with molecular phylogenetic verification of the results, which required fresh material. The authors searched for an older name and hypothesized that their putative new species belonged to either U. tylorhiza (Nyl.) Nyl. (Davydov Reference Davydov, Andreev and Himelbrant2017; Davydov et al. Reference Davydov, Peršoh and Rambold2017) or U. koidzumii M. Satô (Davydov et al. Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayer2020b). They also examined the types of U. vellea, U. koidzumii, U. cinereorufescens and the original material of U. cirrhosa Hoffm.
Although the putative new taxon was initially identified through molecular phylogenetic analysis, the first author later detected it in herbarium specimens. During his time at Helsinki University's herbarium (H), ED discovered old specimens with distinct diagnostic traits that were collected in the vicinity of Helsinki in the 1930s. With the kind assistance of Professor Teuvo Ahti, we located this site and examined, in detail, a rock outcrop containing hundreds of specimens of this species. We have designated this collection as the type and named it as a new species in honour of Teuvo Ahti.
A final decision to describe the species as new could not be made until the status of Umbilicaria koidzumii was confirmed. YO attempted to collect this species in the locus classicus several times before succeeding. The obtained sequences and examined traits suggest that U. koidzumii is related to U. cinereorufescens and is distributed worldwide, while U. cinereorufescens s. str. may be endemic to Europe. As a result, the status and circumscription of two well-known species of the U. vellea group, namely U. vellea and U. cinereorufescens, are changing significantly. Although the data currently available are insufficient to develop a monograph for the entire group, we aim here to define the new species in relation to the U. vellea group and to discuss other potentially confusing taxa, such as U. koidzumii, U. cinereorufescens and U. trabeculata Frey & Poelt, while also resolving related nomenclature issues.
Materials and Methods
Sampling
Fresh material of the putative new species, U. cinereorufescens s. lat. and U. vellea was collected by ED. Specimens of U. koidzumii were collected by YO in the type locality in Japan. Type and other materials of the U. vellea group were examined from the herbaria ALTB, CHR, G, GZU, H, HMAS, KPABG, LE, LECB, M, MAG, MW, O, OSC, TNS and UHU, and the personal collections of S. V. Chesnokov, L. S. Konoreva, B. McCune, D. Masson and G. P. Urbanavicus. A high resolution stacked image of the rhizinomorphs of the lectotype of U. vellea (LINN) was provided by Dr Mark Spencer (LINN). We used sequences obtained from GenBank exclusively from specimens we had previously studied morphologically, anatomically and chemically. Details of the material and GenBank Accession numbers are presented in Table 1. We included two distinct morphotypes of U. cinereorufescens in the analyses. The morphotype of ‘U. cinereorufescens-1’ collected in Norway corresponds with the type in having very irregular, tuberose to peg-like, often flattened rhizinomorphs. In contrast, ‘U. cinereorufescens-2’ has predominantly peg-like, ball-tipped rhizinomorphs, usually with additional black filiform rhizinomorphs.
Table 1. Voucher information and GenBank Accession numbers for the specimens used in the phylogenetic analyses. New sequences generated in this study are in bold.
Morphology and anatomy
Morphological observations were carried out using a stereomicroscope (Zeiss Stemi 2000-C). Apothecia and thalli were hand-sectioned with a razor blade and observed in water mounts with a compound microscope (Zeiss Axio Lab.A1.). The images were captured using a Zeiss AxioCam ERc5s (resolution 5 Mpx) colour digital camera with the software ZEN 2012 (Carl Zeiss). Measurements were recorded to the nearest 0.5 μm. Measurements of ascospores and thalloconidia are presented as follows: (smallest value recorded–) (x̄ – SD) – x̄ – (x̄ + SD) (–largest value recorded), where x̄ is the (arithmetic) sample mean, and SD the sample standard deviation. The two extreme values and the sample mean are given to the nearest 0.5 μm and n = the number of measurements made. Other measurements are presented as (extreme minimum) minimum–maximum (extreme maximum).
Chemical analyses
Lichen substances were studied by spot tests using potassium hydroxide solution (K), sodium hypochlorite solution (C), 1,4-p-phenylendiamine (PD), and by thin-layer chromatography (TLC) with solvent systems A, B' and C following Orange et al. (Reference Orange, James and White2001). High performance liquid chromatography (HPLC) analysis was performed as described in Davydov et al. (Reference Davydov, Blum, Kashevarov and Grakhov2019).
DNA extraction, amplification and sequencing
Single thallus parts (c. 5 mm2) or 3–4 apothecia were carefully checked for fungal infections and thoroughly cleaned of extraneous matter. DNA extraction, amplification and sequencing followed the methods of Davydov & Yakovchenko (Reference Davydov and Yakovchenko2017) and Ohmura et al. (Reference Ohmura, Sugimoto, Yakovchenko and Davydov2022). To test the phylogenetic relationships within the Umbilicaria vellea group, the internal transcribed spacer region of nuclear ribosomal DNA (ITS), the large subunit of the mitochondrial ribosomal DNA (mtLSU), and RNA polymerase II between six and seven conserved parts (RPB2) were amplified in a single reaction from DNA extracts. The primers used in this study are shown in Table 2. The program Geneious v. 6.0 (Biomatters Ltd, New Zealand) was used for assembling partial and complementary sequences.
Table 2. Summary statistics, PCR settings and substitution models used for the different datasets.
Sequence alignment and phylogenetic analyses
New sequences of species of the U. vellea group were supplemented with sequences obtained during a study of Umbilicariaceae phylogeny (Davydov et al. Reference Davydov, Peršoh and Rambold2017, Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayer2020b; Davydov & Masson Reference Davydov and Masson2022), representing different subgenera with an emphasis on Umbilicaria subg. Papillophora; Xylopsora friesii (Ach.) Bendiksby & Timdal was used as the outgroup. This selection is based on the studies of Wedin et al. (Reference Wedin, Wiklund, Crewe, Döring, Ekman, Nyberg, Schmitt and Lumbsch2005), Bendiksby & Timdal (Reference Bendiksby and Timdal2013) and Davydov et al. (Reference Davydov, Peršoh and Rambold2017), in which Xylopsora Bendiksby & Timdal forms the sister clade to Umbilicaria. GenBank Accession numbers are provided in Table 1.
Three single-gene datasets containing the sequences listed in Table 1 were compiled and aligned using the MAFFT algorithm (Katoh et al. Reference Katoh, Rozewicki and Yamada2019). Introns in the mtLSU sequences were manually removed from alignments. All U. cinereorufescens and U. vellea mtLSU sequences contained introns, whereas none of U. ahtii and seven of 12 U. koidzumii sequences contained introns at the same position. Before combining sequences into a joint ITS + mtLSU + RPB2 data matrix, the unambiguously alignable regions were used to calculate single-marker phylograms (not shown) using the online version of IQ-TREE (Nguyen et al. Reference Nguyen, Schmidt, von Haeseler and Minh2015; Trifinopoulos et al. Reference Trifinopoulos, Nguyen, von Haeseler and Minh2016), and these were tested for conflicts among datasets. Since the cladograms were similar regarding well-supported (BS ≥ 70%) clades and therefore lacked conflicts, sequences were combined into one matrix consisting of 2030 sites and used for maximum likelihood and Bayesian analyses.
A heuristic search for the maximum likelihood (ML) bootstrap tree with simultaneous inference of the optimal partitioning scheme and substitution models for each data partition was performed using the online version of IQ-TREE (Nguyen et al. Reference Nguyen, Schmidt, von Haeseler and Minh2015; Trifinopoulos et al. Reference Trifinopoulos, Nguyen, von Haeseler and Minh2016), suggesting seven initial partitions (ITS1; 5.8S rDNA; ITS2; mtLSU; RPB2 1st, 2nd, and 3rd codon positions). Branch lengths were assumed to be equal for all partitions. Branch support was estimated with the ultrafast bootstrap algorithm (Minh et al. Reference Minh, Nguyen and von Haeseler2013) based on 1000 bootstrap replicates and using a maximum of 1000 iterations and a minimum correlation coefficient of 0.99 as a stopping rule.
To provide additional support for our phylogenetic reconstruction, we ran a Bayesian analysis. We used the Markov chain Monte Carlo (BMCMC) method (Larget & Simon Reference Larget and Simon1999) implemented in MrBayes v. 3.2.3 (Ronquist et al. Reference Ronquist, Teslenko, van der Mark, Ayres, Darling, Höhna, Larget, Liu, Suchard and Huelsenbeck2012) to infer phylogenetic trees, applying the partitioning scheme inferred with IQ-TREE and slightly simplified substitution models inferred by PartitionFinder v. 1.1.1 (Lanfear et al. Reference Lanfear, Calcott, Ho and Guindon2012), because most of the models inferred by IQ-TREE are not implemented in MrBayes. Three parallel analyses, each with six incrementally heated chains using the default heating factor of 0.2, were run for 100 million generations; every 200th generation was sampled until the average standard deviation (ASD) of split frequencies had dropped to 0.01. This was the case after 78 M generations. The first 50% of trees was discarded as burn-in and a 50% majority-rule consensus tree was calculated from the remaining trees of the three runs using the sumt command implemented in MrBayes v. 3.2.3.
Results
The phylogenetic study
For the phylogenetic analyses, we used a total of 208 sequences, 77 of which were obtained during this study and 131 in previous investigations (Davydov et al. Reference Davydov, Peršoh and Rambold2017, Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayer2020b; Davydov & Masson Reference Davydov and Masson2022) and deposited in GenBank (Table 1). To test the monophyly and phylogenetic relationships of the species, we used three markers as single gene matrices and in combined datasets. Summary statistics are shown in Table 2. ITS and mtLSU sequences were successfully obtained from a total of 18 specimens, including 15 specimens of Umbilicaria ahtii and three specimens of U. koidzumii from Japan. In addition, RPB2 sequences were obtained from three specimens of each of the species, U. ahtii and U. koidzumii. Although the set of sequences is incomplete with respect to the phylogenetic markers (Table 1), ITS or mtLSU markers were obtained for all specimens. The phylograms based on ITS or the more conservative marker mtLSU (Supplementary Material Figs S1 & S2, available online) contained similar, well-supported lineages for separate species or their regional populations. The RPB2 phylogram (Supplementary Material Fig. S3, available online) included only a small fraction of specimens, and the relationships in the backbone were not statistically supported. An ITS + mtLSU phylogram (Supplementary Material Fig. S4, available online), in which both phylogenetic markers were obtained for each specimen, had a similar topology to the ITS + mtLSU + RPB2 phylogram, in which 1–3 phylogenetic markers were obtained for each specimen, and both phylograms provided high support for most of the clades. Therefore, we present and discuss the ITS + mtLSU + RPB2 phylogram (Fig. 1), which includes specimens from a wider geographical range. The combination of the phylograms from the Bayesian 50% majority-rule consensus tree and the IQ-tree (Fig. 1) share the same topology.
Figure 1. Phylogenetic relationships amongst Umbilicaria species, based on a maximum likelihood analysis (IQ) of ITS1-5.8S-ITS2, mtLSU and RPB2. Numbers at tree nodes indicate bootstrap values of ML (left) and BMCMC posterior probabilities (right). Thicker branches indicate when both the bootstrap value of ML is ≥ 70% and the BMCMC posterior probability is ≥ 0.95. Branch lengths represent the estimated number of substitutions per site assuming the respective models of substitution (Table 2). The basal branches of Umbilicaria are collapsed to reduce the overall figure size. ‘U. cinereorufescens-1’ and ‘U. cinereorufescens-2’ represent two distinct morphotypes of U. cinereorufescens, and sequences of typical U. koidzumii from the locus classicus are indicated. Ru = Russia, P = Peninsula. In colour online.
Both Umbilicaria ahtii and U. koidzumii were grouped within Umbilicaria subg. Papillophora. All specimens of U. ahtii formed a well-supported clade for all single-marker (Supplementary Material Figs S1–S3) and combined phylograms (Fig. 1, Supplementary Material Fig. S4). Umbilicaria ahtii was grouped as sister to U. meizospora, while the morphologically similar U. vellea was grouped as sister to U. hirsuta + U. josiae Frey. Sequences of U. ahtii were nearly identical and no geographical segregation was observed, except for two sequences from Alaska which formed a well-supported clade nested within the remaining sequences. Umbilicaria vellea from the northern Far East (Chukotka, Koryakia and North Kurile) clustered as sister to the remaining U. vellea, but without support (76% BS; 0.7 PP).
Sequences of Umbilicaria cinereorufescens formed two well-supported groups (1 and 2; Fig. 1). Sequences of typical U. koidzumii from the locus classicus nested within the ‘U. cinereorufescens-2’ clade. The ITS sequences of the Japanese material were identical, except for the fertile U. koidzumii 7 (Supplementary Material Fig. S1).
Morphology, anatomy and secondary chemistry
Upon careful examination and comparison of the sequenced material of Umbilicaria vellea and the putative new species, described below as U. ahtii (Fig. 2), we found that rhizinomorphs, apothecia and ascospores are diagnostic traits for these species (refer to Table 3, Figs 3 & 4). However, apothecia and ascospores are not useful for routine identification since apothecia develop infrequently and well-developed spores are rarely found.
Figure 2. Habitat and morphology of Umbilicaria ahtii. A, type locality of U. ahtii: Helsinki, Lankiniemi near the bridge to Tarvo, rock outcrops covered by pine forest. B, general habit in the field (type locality). C, schizidia (develop occasionally, ALTB L5418). D & E, gyrodisc apothecia with a visible crack, a black lower cortex and a ‘corona’ of rhizinomorphs (ALTB L6939). F & G, general habit of the holotype (LE-L26006). F, upper side. G, lower side. Scales: C–E = 1 mm; F & G = 1 cm. In colour online.
Table 3. Diagnostic traits of selected species of Umbilicaria vellea group. M = major compound, m = minor compound, tr = trace.
Figure 3. Umbilicaria ahtii (ALTB L5747). A–D, rhizinomorphs. E, section of thallus. F, section of apothecium. G, simple and submuriform ascospores. Scales: A–D = 1 mm; E & F = 100 μm; G = 10 μm. In colour online.
Figure 4. A & B, Umbilicaria ahtii (A, KPABG L4482; B, Melekhin ki433 (ALTB)), lower surface, longer light and shorter black, branched rhizinomorphs with bundles of thalloconidia. C, Lichen velleus (LINN, lectotype), the lower surface with tuberose and vermiform rhizinomorphs covered by thalloconidia. D, Umbilicaria vellea (Davydov 16229; ALTB), light-coloured lower surface with clusters of black thalloconidia. E, Umbilicaria vellea (ALTB L6095), vermiform rhizinomorphs with basal parts covered by cases of thalloconidia. F, Umbilicaria vellea (Davydov 6535; ALTB), light brown lower surface with longer light-coloured rhizinomorphs and black tuberose types with thalloconidia. Scales = 1 mm. In colour online.
Previous authors have adhered to a broad concept of Umbilicaria cinereorufescens, which included specimens with a black areolate lower surface and black thalloconidial rhizinomorphs of different shapes, including filiform and irregular, sometimes flattened, tuberose to peg-like, short and long, often ball-tipped. The type material of U. cinereorufescens lacks filiform and long, ball-tipped rhizinomorphs. However, some other specimens from Europe, as well as the majority of investigated specimens from Asia and North America, do have them. Therefore, we assume that the varying shape of black rhizinomorphs is a characteristic trait of U. cinereorufescens. Umbilicaria vellea can be easily distinguished from all those specimens by its thin, beige-coloured, comparatively long rhizinomorphs, in addition to shorter, peg-like and black rhizinomorphs covered by thalloconidia. The present phylogenetic analyses revealed that U. cinereorufescens is represented by two clades, namely ‘U. cinereorufescens-1’ and ‘U. cinereorufescens-2’ (Fig. 1). All specimens of ‘U. cinereorufescens-1’ were collected from various locations in Norway and included morphotypes corresponding to the type material of U. cinereorufescens (Fig. 5A & B). Specimens of ‘U. cinereorufescens-2’ possess black filiform rhizinomorphs in addition to the predominate peg-like ball-tipped rhizinomorphs (Fig. 5F–I), which are sometimes segregated in tiers of shorter and longer lengths. Although ‘U. cinereorufescens-1’ and ‘U. cinereorufescens-2’ produce the same secondary compounds, such as gyrophoric, lecanoric and crustinic acids, there is a quantitative difference between them. Umbilicaria ‘cinereorufescens-1’ consistently contains crustinic acid as a major compound, while ‘U. cinereorufescens-2’ contains gyrophoric acid as a major compound, with crustinic acid sometimes present in trace amounts (Table 3).
Figure 5. A & B, lectotype of Umbilicaria cinereorufescens (G 53127). A, upper side. B, lower side with black irregular tuberose rhizinomorphs covered by thalloconidia. C–E, holotype of Umbilicaria koidzumii (TNS L119408). C, upper side. D, lower side with two types of rhizinomorphs. E, lower side with longer light and shorter black ball-tipped rhizinomorphs. F–I, lower side of Umbilicaria koidzumii with black ball-tipped rhizinomorphs. F, Alaska, USA (McCune 31407). G, Altai, Russia (Davydov 651). H, New Zealand (Galloway 8095). I, Japan (Ohmura 13921). Scales: A & C = 1 cm; B & D–I = 1 mm. In colour online.
The holotype of Umbilicaria koidzumii (Fig. 5C–E) does not quite fit the circumscription of U. cinereorufescens because, similar to U. vellea specimens, it has two types of rhizinomorphs: shorter and black with multicellular thalloconidia, and longer and pale. However, unlike U. vellea and similar to ‘U. cinereorufescens-2’, the light-coloured rhizinomorphs often have balls of thalloconidia at the tips and the species also has submuriform ascospores. Of the five sequenced specimens collected around the type locality of U. koidzumii, two fully corresponded to the holotype (U. koidzumii 12 & 13). These two specimens had light and long rhizinomorphs with balls of thalloconidia at their tips. One specimen (U. koidzumii 7) was fertile. This fertile specimen, along with the two remaining specimens (U. koidzumii 14 & 15), fit our conception of U. cinereorufescens s. lat. Therefore, we can observe both black and light-coloured ball-tipped rhizinomorphs in the same collections from Japan. The morphological variation of specimens with identical or very similar ITS, mtLSU and RPB2 sequences allowed us to better morphologically circumscribe U. koidzumii. Therefore, it could be applied to the entire ‘U. cinereorufescens-2’ clade since it contains sequences of specimens that share the same traits but differ from ‘U. cinereorufescens-1’. We have summarized the diagnostic traits of four species in Table 3. After considering anatomical, morphological, chemical and phylogenetic data, there is sufficient evidence to describe a new species and amend the circumscriptions of U. cinereorufescens, U. koidzumii and U. vellea.
Taxonomy
Umbilicaria ahtii Davydov sp. nov.
MycoBank No.: MB 854205
Resembles Umbilicaria vellea but with a darker lower surface, dark brown to black longer rhizinomorphs, and lacking thalloconidia directly on the lower surface and basal part of the rhizinomorphs. Multicellular thalloconidia develop exclusively on the tips and branches of short or long unbranched or branched rhizinomorphs. Ascospores are larger and submuriform.
Type: Finland, Uusimaa, Helsinki, Munkkiniemi, Lankiniemi near the bridge to Tarvo, 60°12′10.04″N, 24°51′07.98″E, elev. 5 m, rock outcrops covered by pine forest, W-exposed wet steep granite rock facing the sea bay, 27 October 2018, E. A. Davydov 17823 (LE L-26006—holotype; ALTB, FR, GZU, H, O, TNS—isotypes). GenBank Accession nos: PP849072, PP849105.
Thallus monophyllous, frequently of medium size, 3–6(–18) cm diam. and 0.2–0.4 mm thick, umbilicate, rigid, undulating with broad folds, the margins first entire, later incised and often wrapped up or downwards; upper surface dull, minutely rimose to areolate, often pruinose; pale to dark grey, with a brown or violet tinge, senescent thalli getting brown, occasionally with rhizinomorphs protruding through cracks or splits; lower surface of juvenile thalli light brown, later entirely black or lighter towards the margins, smooth to rimose or finely areolate with abundant rhizinomorphs of several types in varied amounts; trabeculae emerging from the umbilicus often present and can also produce rhizinomorphs. Two main classes of rhizinomorphs occur, non-thalloconidial and thalloconidial; non-thalloconidial rhizinomorphs long, 1–2(–4) mm, grey-brown to black, tree-like branched (Fig. 3D), with one main thick axis being few or many times iso- or anisotomously branched. Thalloconidial rhizinomorphs are shorter and may be of several types: 1) pin-like, with a usually unbranched axis; 2) ball-tipped and branched, with branched axis 0.5–1 mm and balls of thalloconidia at tips (Fig. 3A); 3) branched rhizinomorphs with thalloconidial bundles (Fig. 3B & C). The thalloconidia never cover the lower surface or base of long rhizinomorphs, but always develop on branches. Vegetative propagules are usually absent. Exceptionally, schizidia (Fig. 2C) can develop under humid conditions. Thalloconidia present, developing exclusively on rhizinomorphs; multicellular with c. 10–50 cells, (15.0–)23.0–28.5–33.5(–42.5) × (10.0–)16.0–20.0–24.0(–30.0) μm (n = 60). Upper cortex paraplectenchymatous, outer part brownish and 10–13 μm thick, inner part hyaline and 20–30 μm thick; epicortex 5–10 μm thick; algal layer continuous, 100–120 μm thick, algae trebouxioid; medulla colourless, dense, 70–125 μm thick, arachnoid plectenchymatous of highly branched pachydermatous hyphae (of Vellea-type (Valladares & Sancho Reference Valladares and Sancho1995)); lower cortex not clearly differentiated from medulla, scleroplectenchymatous, including a colourless inner layer c. 20–25 μm thick, and a brown outer layer 10–20 μm thick (Fig. 3E).
Apothecia rare, occurring at the periphery of thalli, 1–3 mm diam., sessile, gyrose, black; apothecial margin thin, c. 100–150 μm thick. Young apothecia flattened, sunken into depression in the thallus; mature apothecia adnate by a wide base to the lower cortex and separated from the other thallus layers by a crack, so the black surface of the lower cortex, and sometimes also a ‘corona’ of rhizinomorphs, is visible from above (Fig. 2D & E); flattened, overmature apothecia become convex, circular or angular; epihymenium brown, 12.5–17.5 μm thick; hymenium hyaline, 65–95 μm thick; hypothecium light brown, 45–60 μm thick; excipulum in inner part yellowish, in outer part dark brown; paraphyses septate, branched, 1.8 μm thick, sometimes slightly thickened at the tips, up to 1.8–2.0 μm; asci 8-spored; ascospores hyaline, simple to submuriform (11.0–)12.5–14.0–15.0(–17.5) × (6.5–)7.0–8.0–9.0(–10.5) μm (n = 20).
Pycnidia sometimes developing at the periphery of the thalli, 120–150 μm diam., with black prominent ostiole; wall brown, 10 μm thick; pycnoconidia bacilliform, 3.5–4.5 × 1 μm.
Chemistry
Thallus K−, C+ red, KC+ red, Pd−, UV−; all specimens examined contain gyrophoric acid as a major compound and lecanoric acid as a minor compound, with crustinic acid sometimes occurring as a trace detectable by TLC. The holotype contains gyrophoric and lecanoric acids, and a trace of crustinic acid detectable only by HPLC.
Etymology
The species is named after the renowned lichenologist Teuvo Ahti, with huge gratitude for his kindness, responsiveness, care and help, and in recognition of his great contributions to lichenology.
Ecology
Umbilicaria ahtii grows mostly on steep siliceous rocks both near the sea and in the upper mountain belt at an elevation range of 5–1900 m in Fennoscandia and Alaska, and 1500–1900 m in South Siberia. It grows in damp places, often by the sea, lakes and rivers, and in shady places in forests.
Distribution
Umbilicaria ahtii is a Holarctic species currently known from a wide range of localities in Europe, Asia and North America, at latitudes from the High Arctic (Svalbard) to South Siberia.
Notes
Umbilicaria ahtii most resembles U. vellea although these are not sister species. This homoplasy can be explained by the same environmental influences on these species growing in similar conditions. The description and photographs in Poelt & Nash (Reference Poelt and Nash1993) are of U. vellea s. str. Umbilicaria ahtii has a Holarctic distribution. After distinguishing U. ahtii and U. koidzumii, the true distribution of U. vellea should be revised.
The rhizinomorph features are essential in distinguishing U. ahtii based on morphology. However, the separation of rhizinomorphs into ‘non-thalloconidial’ and ‘thalloconidial’ is not absolute, because sometimes a small number of balls of thalloconidia may develop at some distal or basal branches of ‘non-thalloconidial’ rhizinomorphs.
Branching of basically non-thalloconidial rhizinomorphs is mainly at an acute angle (c. 45° to 60°) and not closely intertwined, giving the underside of the thallus a hairy appearance due to the many long ±parallel branches. This type develops on juvenile thalli first. Branched rhizinomorphs with thalloconidial bundles may be individual to well developed, covering almost the whole lower surface under the longer non-thalloconidial rhizinomorphs. Also, branched rhizinomorphs with thalloconidial bundles can be in two size classes on the same thallus, so that rhizinomorphs segregate into three tiers: longer and middle-size branched and shorter pin-like. In addition, a secondary kind of thalloconidial rhizinomorph may be present: short, 0.2–0.4 mm long, radiating at the ends into thin branches bearing numerous thalloconidia, with a coral-like appearance. Over time, the axes may become indistinguishable due to the mass of thalloconidia.
Umbilicaria cinereorufescens (Schaer.) Frey
Hedwigia 71, 109 (1931).—Umbilicaria vellea γ spadochroa e. cinereorufescens Schaer., Enum. Crit. Lich. Eur., 25 (1850).—Gyrophora cinereorufescens (Frey) Schol., Nytt Mag. Naturvidensk. 75, 28 (1934); type: [Switzerland] Grimsel, Schaerer, in Lich. Helv. Exs. no. 142 (G 53127!—lectotype designated by Hestmark (Reference Hestmark2007)).
The circumscription and species status of Umbilicaria cinereorufescens was proposed by Frey (Reference Frey1929, Reference Frey1931) who examined herbarium material and collected specimens in the Alps, including from Grimsel, and from Dovre in Norway (Frey Reference Frey1929). All our sequences of ‘U. cinereorufescens-1’ are from Norway and correspond to the lectotype and the description by Frey (Reference Frey1931). Regarding the rhizinomorphs, Frey (Reference Frey1931: 110) mentioned that ‘…fädige Rhizinen fehlen konstant.’ [filiform rhizines are constantly absent]. This trait seems to be diagnostic for U. cinereorufescens s. str. The distribution of U. cinereorufescens s. str. is poorly known due to previous misunderstandings. Umbilicaria cinereorufescens s. str. is known from the Alps and Scandinavia, and probably restricted to Europe or West Eurasia. The endemic status seems plausible in that West Eurasia is one of the centres of speciation and endemism in Umbilicaria subg. Papillophora (Frey Reference Frey and Zahlbruckner1933; Davydov Reference Davydov2022; Davydov & Masson Reference Davydov and Masson2022).
Umbilicaria cirrhosa Hoffm.
Descr. Adumb. Plant. Lich. 1(1), 9 (1790) as ‘cirrosa’; type: Icon.: Descr. Adumb. Plant. Lich. 1(1), 9 (1789) [1790]: Tab. II, figs 3 & 4.
Umbilicaria koidzumii Yasuda ex M. Satô
J. Jap. Bot. 11, 314 (1935); type: Japan, Honshu, Prov. Kai, Mt Komagatake, 27 July 1921, H. Koidzumi (TNS L119408—holotype!; H—isotype!).
Umbilicaria koidzumii was described as a species related to U. esculenta (Miyoshi) Minks (Satô Reference Satô1935) and was thought to be endemic to Japan (Ohmura & Kashiwadani Reference Ohmura and Kashiwadani2018), based on the type specimens and several additional collections (TNS) identified by ED. The holotype of Umbilicaria koidzumii resembles U. vellea and U. cinereorufescens rather than U. esculenta as suggested by Satô (Reference Satô1935). The original description of U. koidzumii (Satô Reference Satô1935) was based on senescent specimens. The combination of short pin-like and long ball-tipped branched rhizinomorphs led Davydov & Masson (Reference Davydov and Masson2022) to suggest that U. koidzumii is probably conspecific with ‘U. cf. koidzumii from Kola Peninsula’ (=U. ahtii). The presence of light-coloured rhizinomorphs is a rare and probably atavistic trait in U. koidzumii; it has been observed in large, probably senescent specimens from Japan and New Zealand.
According to the current molecular phylogenetic data, U. koidzumii is not an endemic species of Japan; the species is distributed in the Holarctic (Europe, Asia and North America) and in the Southern Hemisphere (New Zealand). Sequence data suggest that U. koidzumii also occurs in South America. We have tested sequences of U. vellea and U. cinereorufescens from GenBank but have not included them in the phylogram because we have not examined the corresponding specimens. Sequences of both U. cinereorufescens (GenBank Accession numbers HM161503 and HM161511) and U. vellea (HM161490) from equatorial South America clustered within U. koidzumii. The identification as U. vellea (Hestmark Reference Hestmark2016) suggests that the specimen probably also has slender rhizinomorphs, as we have mentioned for specimens from Japan and New Zealand.
Umbilicaria trabeculata is another species which may be related to U. koidzumii. In describing Umbilicaria trabeculata, Poelt (Reference Poelt1977) emphasized the unusual nature of its rhizinomorphs as ‘strongly warty in appearance …, often twisted in a wavy manner, … occasionally [with] the compacted ends come together to form a coarse-blackish second surface.’ Hestmark (Reference Hestmark1990) had not studied the type but suggested that it might be a synonym of U. cinereorufescens, which also has identical multicellular thalloconidia. Poelt & Nash (Reference Poelt and Nash1993) accepted U. trabeculata as a separate species and Wei & Jiang (Reference Wei and Jiang1993) synonymized it with U. tylorhiza from the Kola Peninsula. This concept was followed by Davydov et al. (Reference Davydov, Peršoh and Rambold2017), who named a specimen from Yunnan as U. tylorhiza. It is noteworthy that Poelt & Nash (Reference Poelt and Nash1993) mentioned that the original description of U. trabeculata from the Himalayas (Poelt Reference Poelt1977) was based on rather sparse material, but examination of additional material from Asia, Africa and North America changed Poelt's opinion on the diagnostic characters of U. trabeculata, which should include ‘partly filiform, partly irregular, often ball-tipped rhizinomorphs’. This trait fully corresponds to our current concept of U. koidzumii. Most of the specimens identified by Poelt as U. trabeculata from GZU were re-identified by the first author as U. cinereorufescens (s. lat.) or ‘U. vellea-2’, the former actually belonging to U. koidzumii. Umbilicaria trabeculata differs from both U. cinereorufescens and U. koidzumii mainly by the characteristics of the rhizinomorphs, but this trait is variable and more specimens with a range of morphologies are needed for it to be conclusive. At present we do not have enough material to conclude whether U. trabeculata is a separate species or not.
Umbilicaria tylorhiza (Nyl.) Nyl.
Flora, Regensburg 52, 389 (1869).—Umbilicaria vellea (L.) Ach. [subsp.?] tylorhiza Nyl., Not. Sällsk. Fauna et Fl. Fenn. Förh., Ny Ser. 8, 122 (1866).—Gyrophora tylorhiza (Nyl.) Nyl. in Hue, Rev. Bot. 5, 14 (1887) [1886–87]; type: [Russia, Murmansk Region], Lapponia orientalis, Kantalahti, 1861, N. I. Fellman (H-NYL 31524 = H 9510903—lectotype designated here!, MycoBank Typification no.: MBT 10020878; H 9503606—isolectotype!).
Molecular phylogenetic data support a Holarctic distribution of U. vellea (Fig. 1, Table 1), while the presence of the species in the Southern Hemisphere requires further investigation. A worldwide revision of the Umbilicaria vellea and U. cinereorufescens groups is needed before we can confidently characterize the distribution and ecology of the species.
Umbilicaria vellea (L.) Michx.
Fl. Boreali-Americ. (Paris) 2, 323 (1803), nom. illegit., ICN (Shenzhen Code) Art. 53.1.—Lichen velleus L., Sp. Pl. 2, 1150 (1753).—Gyrophora vellea (L.) Ach., Methodus, Sectio prior (Stockholmiæ), 109 (1803); type: “L[ichen] velleus” (LINN-HL 1273-199 (high resolution digital photograph!, Fig. 4C), lower left specimen—lectotype, designated here, MycoBank Typification no.: MBT 10020613).
Nomenclatural notes
Lichen velleus
Howe (Reference Howe1912: 201) designated the entire sheet LINN-HL 1273-199, annotated by Linnaeus, as the type (lectotype) of Lichen velleus L. Since the sheet contains three specimens, the lectotype was restricted to the right-hand (largest) specimen by Jørgensen et al. (Reference Jørgensen, James and Jarvis1994: 364), who did not explain that choice. We do not consider this choice optimal, since this specimen is glued to the sheet of paper and we can see only its upper surface, which lacks apothecia and looks like many species of the U. vellea group. However, diagnostic traits (i.e. rhizinomorphs and thalloconidia) can be examined on the only specimen that is glued upside down, so that both wrapped edges of the upper surface and lower surface are available for examination. This specimen is therefore designated as the lectotype. The high-resolution digital photograph of the lower surface showed characteristic light-coloured, longer rhizinomorphs as well as tuberose and vermiform types, covered by thalloconidia (Fig. 4C).
Umbilicaria cirrhosa
The figures of Umbilicaria cirrhosa in Hoffmann (Reference Hoffmann1790) or in Dillenius (Reference Dillenius1768) lack details of the rhizinomorphs which could help to adequately identify species of the U. vellea group in its modern concept. The only specimen of U. cirrhosa was traced in Hoffmann's General Herbarium (Fig. 6C–F). The specimen at MW belongs to the personal collection of Hoffmann, which was purchased from him by Moscow University (Hoffmann Reference Hoffmann1825; Sokoloff et al. Reference Sokoloff, Balandin, Gubanov, Jarvis, Majorov and Simonov2002). This specimen was certainly examined by Hoffmann. The characteristics of this specimen do match well with the original description, but it evidently also belongs to U. vellea. The label reads ‘cirrhosa’, whereas the protologue reads ‘cirrosa’ (Hoffmann Reference Hoffmann1790: 9). Vainio (Reference Vainio1888) examined the Hoffmann's Herbarium specimen in MW and identified it as ‘Gyrophora spodochroa (Ehrh.) Ach.’, but it clearly belongs to U. vellea in its modern concept (Fig. 6C).
Figure 6. A & B, isolectotype of Umbilicaria tylorhiza (H 9503606). A, upper side. B, lower side. C–F, Umbilicaria cirrhosa (MW-Hoffm 8603a). C, upper side. D, lower side. E, label written in Hoffmann's handwriting. F, label from MW. Scales: A–D = 1 cm. In colour online.
Umbilicaria tylorhiza
Nylander (Reference Nylander1866) cited the collection ‘Kantalahti (N. I. Fellman 1861)’ in the protologue. Two duplicate specimens were traced in H herbarium (H 9510903 and H 9503606); the first, having the label written in Nylander's handwriting and including a positive C reaction mark (H-NYL 31524), was cited as ‘holotype’ by Wei & Jiang (Reference Wei and Jiang1993) and designated here as the lectotype of the name U. tylorhiza. The first author examined both specimens several times and finally came to the conclusion that it is a senescent specimen of U. vellea. This is supported by the following features: the brown, finely rimose lower surface and the presence of slender rhizinomorphs (some of which are broken off) in addition to the capitate rhizinomorphs which are common to many species. The predominance of capitate thalloconidial rhizinomorphs and the degradation of the slender ones suggest that this senescent specimen probably grew in moist conditions. Thalloconidia may develop around the base of simple slender rhizinomorphs, as is characteristic for U. vellea but not for other species of the U. vellea group with similar multicellular thalloconidia, (i.e. U. ahtii, U. cinereorufescens, U. koidzumii, and U. trabeculata), which also, in contrast to U. vellea, have a lower surface that soon turns dark and rough areolate. Extensive material of both U. ahtii and U. vellea from Kantalahti has been examined at KPABG, and we have also obtained sequences of both species from the locus classicus of U. tylorhiza.
Umbilicaria vellea
A species of Umbilicaria with the specific epithet ‘vellea’ was first introduced by Hoffmann (Reference Hoffmann1794). He referred the specimen ‘Lichen polyrhizus Lightf. Scot. 864’ from Flora Scotica by Lightfoot (Reference Lightfoot1777) to Umbilicaria vellea. The species description provided by Hoffmann is inadequate to distinguish between the species of Umbilicaria in their current circumscription. However, the figures (Tabula XXVI no. 3) made from Ehrhart's specimens from Uppsala clearly belong to U. polyrhizos (L.) Stenh. in the modern sense. The polyphyllous thallus, brown upper surface, entirely black lower surface, and especially the apothecia with radial gyri (i.e. of actinodisc type (Scholander Reference Scholander1934)) are diagnostic traits for U. polyrhizos. In the protologue of U. vellea, Hoffmann (Reference Hoffmann1794) also referred to Linnaeus’ Lichen velleus, but with a question mark, indicating that he was not sure whether L. velleus belonged to U. vellea or not. We have recognized the specimen of U. vellea in the lichen collection of Hoffmann in MW; specimen no. 8603 in Hoffmann's Herbarium in MW belongs to U. polyrhizos and could be a candidate for the lectotype of the name U. vellea Hoffm.
Umbilicaria vellea Hoffm. is, therefore, a synonym of U. polyrhizos. Combinations of U. vellea with Lichen velleus L. as basionym proposed by subsequent authors are later homonyms and therefore illegitimate under ICN (Shenzhen Code) Art. 53.1. At present, the legitimate name for the species which we currently refer to as ‘U. vellea’ is Umbilicaria cirrhosa Hoffm.
In order to achieve nomenclatural stability, we propose to conserve the well-known but homonymous name Umbilicaria vellea (L.) Michx., historically applied to a widespread lichen species known at least since Linnaeus (Reference Linnaeus1753), against Umbilicaria vellea Hoffm. Such a proposal is in preparation.
Discussion
Phylogenetic position
The phylogenetic position of Umbilicaria ahtii outside of U. vellea was initially assumed based on previous molecular phylogenetic analyses (Davydov et al. Reference Davydov, Peršoh and Rambold2017; Davydov & Masson Reference Davydov and Masson2022). Umbilicaria ahtii belongs to the U. vellea group and phenotypically fully corresponds to this group by having a medium to large, rigid grey thallus with a smooth to areolate, never pustulate or reticulate upper thallus surface, a lower surface with trabeculae emerging from the umbilicus, the lower surface and rhizinomorphs being characteristically papillose or areolate, and simple to submuriform ascospores.
Umbilicaria ahtii clusters as sister to U. meizospora. Both species have apothecia of very similar appearance, and the senescent thalli of U. ahtii resemble U. meizospora in the brown upper surface and black lower surface with black rhizinomorphs giving a hairy appearance. Umbilicaria meizospora and U. ahtii are another example of divergence by alternation of the reproduction mode, as shown for many groups of lichens including Umbilicaria (Poelt Reference Poelt1970; Hestmark Reference Hestmark1990; Buschbom & Mueller Reference Buschbom and Mueller2006; Davydov et al. Reference Davydov, Peršoh and Rambold2017). Umbilicaria meizospora reproduces sexually and U. ahtii mostly asexually. Both species have very clear diagnostic traits and are well defined and distinct in appearance. Sexually reproducing U. meizospora is remarkably restricted in distribution and is endemic to south-western Europe (Davydov & Masson Reference Davydov and Masson2022).
Our results suggested two alternatives for the taxonomic treatment of Umbilicaria koidzumii within U. cinereorufescens s. lat.: either synonymize U. koidzumii with the previously described U. cinereorufescens, or distinguish U. cinereorufescens s. str. and U. koidzumii as an independent species, thereby identifying all specimens of ‘U. cinereorufescens-2’ as U. koidzumii. We chose the latter option because: 1) diagnostic traits were observed in each species; 2) the genetic distance between the clades is relatively large, possibly indicating a somewhat distant relationship between the species, especially considering that statistical support for the ‘U. cinereorufescens-1’ + ‘U. cinereorufescens-2’ clade is relatively low (85% BS; 0.92 PP).
Umbilicaria vellea is represented by two major clades in the present phylogeny with a clear geographical pattern. We were unable to find any morphological differences between the north-west Asian specimens and those from other parts of the U. vellea range. This segregation may reflect a geographical isolation of north-west Asian populations or it may indicate another unrecognized species within U. vellea. Since no diagnostic traits were found and the north-west Asian clade lacked statistical support, we concluded that it does not require a taxonomic rank.
Diagnostic traits
All species discussed possess rhizinomorphs and thalloconidia and are rarely fertile. The size and septation of thalloconidia have been shown to be highly species-specific in some cases (Poelt Reference Poelt1977; Hasenhüttl & Poelt Reference Hasenhüttl and Poelt1978; Hestmark Reference Hestmark1990; Krzevicka Reference Krzewicka2010; Davydov et al. Reference Davydov, Yakovchenko, Urbanavichene, Konoreva, Chesnokov, Kharpukhaeva and Obermayer2020b). However, this is not the case in the Umbilicaria vellea and U. cinereorufescens groups, in which all species show multicellular thalloconidia of similar size, with a mean between 22–28 × 17–23 μm. The variation in maximum size could be explained by the different number of measurements between species, since we used only our own measurements of thalloconidia for U. ahtii and U. koidzumii, but for U. cinereorufescens and U. vellea we also used literature data (Hestmark Reference Hestmark1990; Poelt & Nash Reference Poelt and Nash1993). However, rhizinomorphs and distribution patterns of thalloconidia on rhizinomorphs are variable and diagnostic (Table 3). Small, non-septate hyaline ascospores are diagnostic for U. vellea. Mature spores of U. ahtii, U. cinereorufescens and U. koidzumii are submuriform. So far, we have observed only hyaline ascospores in U. ahtii. However, we believe they may be brown when mature, like the mature submuriform spores seen in other Umbilicaria species. Although the apothecia of all the species mentioned are similarly gyrose, U. ahtii stands out in that its apothecia are very similar in appearance to those of U. meizospora, the phylogenetically closest species. Mature apothecia of both species are attached to the lower cortex by a wide base and separated from the other thallus layers by a crack, so that the black surface of the lower cortex is visible from above (Fig. 2D & E; Frey Reference Frey and Zahlbruckner1933: 264, fig. 36). However, apothecia and ascospores are not useful diagnostic traits for routine identification of the U. vellea group because they develop too rarely.
All four species discussed contain gyrophoric and lecanoric acids as secondary metabolites, as revealed by TLC. Umbilicaria cinereorufescens s. str. stands out with crustinic acid as the major compound. This chemical feature appears to be rare in the genus Umbilicaria and is shared, as far as is known, only by U. meizospora (Davydov & Masson Reference Davydov and Masson2022) and one chemotype of U. crustulosa (Posner et al. Reference Posner, Feige and Huneck1992; Narui et al. Reference Narui, Culberson, Culberson, Johnson and Shibata1996; Seriña et al. Reference Seriña, Arroyo, Manrique and Sancho1996). This result is in agreement with Posner et al. (Reference Posner, Feige and Huneck1992) and Narui et al. (Reference Narui, Culberson, Culberson, Johnson and Shibata1996) who used European and Himalayan material of U. cinereorufescens for HPLC analyses. It is noteworthy that the major compound of the sister species U. koidzumii is gyrophoric acid, while crustinic acid is also detectable by TLC, although always as a minor compound. Posner et al. (Reference Posner, Feige and Huneck1992) reported hiascic acid (minor) for the isotype of U. koidzumii in addition to gyrophoric (major) and lecanoric (minor) acids, and exactly the same pattern was shown for the U. trabeculata isotype. Similar to U. cinereorufescens, U. meizospora contains crustinic acid as the major compound and the sister species U. ahtii contains only a trace of crustinic acid detectable by HPLC and sometimes by TLC. Umbilicaria vellea sometimes contains umbilicaric acid in addition to gyrophoric and lecanoric acids (Posner et al. Reference Posner, Feige and Huneck1992; Narui et al. Reference Narui, Culberson, Culberson, Johnson and Shibata1996; Seriña et al. Reference Seriña, Arroyo, Manrique and Sancho1996).
Key to species of the Umbilicaria vellea group with thalloconidia
1 1–2(–6)-celled thalloconidia develop on lower surface and rhizinomorphs ……… U. orientalis
6–10 to 30–50-celled thalloconidia develop on rhizinomorphs ……… 2
2(1) Rhizinomorphs of mature specimens vary in size and type and are often segregated in tiers: 1) simple to branched filiform rhizinomorphs without or occasionally with thalloconidia, and 2) thalloconidial rhizinomorphs tuberose to peg-like, branched and ball-tipped, or with thalloconidia produced in bundles ……… 3
Rhizinomorphs only thalloconidial, short, tuberose to knob-like or irregular, never filiform ……… 5
3(2) Peg-like ball-tipped rhizinomorphs predominate (Fig. 5E–I), sometimes segregated in tiers; shorter and longer, black filiform rhizinomorphs usually develop in addition and grow in irregular directions, often from filaments (Fig. 5C–I) ……… ……… U. koidzumii
Peg-like ball-tipped rhizinomorphs rare; rhizinomorphs arranged in two layers; shorter ones bearing thalloconidia and longer ones simple to branched, growing perpendicular to lower surface ……… 4
4(3) Longer rhizinomorphs pale, shorter rhizinomorphs tuberose (Fig. 4F) to peg-like; the bases of simple rhizinomorphs often covered with thalloconidia and worm-like (Fig. 4E) ……… U. vellea
Longer rhizinomorphs dark (Fig. 3D), shorter rhizinomorphs branched with bundles of thalloconidia (Figs 3A–C, 4A & B) and/or peg-like and usually also branched, always with axes, never tuberose or worm-like ……… U. ahtii (Fig. 2)
5(2) Rhizinomorphs abundant, very irregular, tuberose to peg-like, often flattened ……… U. cinereorufescens (Fig. 5A & B)
Rhizinomorphs scarce, separate, short and knob-like ……… U. trabeculata
Acknowledgements
ED and LY are especially grateful to Professor T. Ahti for his many years of support, invaluable help and care during visits to the Botanical Museum at the University of Helsinki, and valuable advice on the nomenclature. ED thanks Dr A. Sennikov (H) and his son for accompanying him on the field excursion to Lankiniemi. We gratefully acknowledge help from the curators and staff of the herbaria cited above. Dr L. Myllys (H), Dr S. V. Chesnokov (LE), Dr D. Galloway (Otago), Dr D. E. Himelbrant (LE), Dr O. A. Kataeva (LE), Dr L. S. Konoreva (LE), Prof. B. McCune (Oregon, USA), Dr A. V. Melekhin (KBABG), Dr D. Masson (Bordeaux), Dr E. Timdal (O), Dr G. P. Urbanavicus (UFU) and Dr I. N. Urbanavichene (LE) are thanked for the herbarium specimens used for this study. We are also grateful to Dr C. Printzen (FR) for the opportunity to sequence a part of the material in his laboratory, and for arranging loans of material to ED; to Dr M. Spencer (LINN) for providing a high-resolution photograph of the rhizinomorphs of the U. vellea lectotype; to Prof. B. McCune (OSC), S. Velmala (H) and the anonymous reviewer for their valuable comments. LY was supported by the state assignment of the Ministry of Science and Higher Education of the Russian Federation (no. 124012400285-7).
Author ORCIDs
Evgeny A. Davydov, 0000-0002-2316-8506; Yoshihito Ohmura, 0000-0003-2557-2761; Lidia S. Yakovchenko, 0000-0002-4342-7771.
Competing Interests
The authors declare none.
Supplementary Material
The Supplementary Material for this article can be found at https://doi.org/10.1017/S0024282924000215.
