Hidden diversity and chemical variation within Parmelia saxatilis group (Lecanorales, lichenized Ascomycota) and description of a new species P. tobolewskiana Authors: Emilia Anna Ossowska, Katarzyna Guzow, Irena Bylińska, Martin Kukwa, Katarzyna Szczepańska, Rafał Szymczyk, Ulf Schiefelbein, Beata Guzow-Krzemińska Published: 2026 ABSTRACT The Parmelia saxatilis group comprises four species in Europe with true isidia that share similar morphology, chemistry, and ecology. These include the widely distributed P. ernstiae, P. saxatilis, and P. serrana, and the potentially endemic species P. rojoi (excluded from the study). They are often referred to as semi-cryptic species, which significantly hinders their proper identification. The present study objectives were (1) to confirm the presence of galbinic acid in Parmelia taxa using the RP-HPLC method; (2) to verify if galbinic acid samples form a clade within the phylogenetic tree; and (3) to provide the first comprehensive chemical profiles of secondary metabolites present in isidiate Parmelia species. The study based on ca. 600 specimens focused on the chemical variability of these three species, paying particular attention to the presence of galbinic acid, rarely reported from this group of lichens. We detected galbinic acid in all studied species, mostly in P. saxatilis, then in P. ernstiae and for the first time in P. serrana. Fourteen chemotypes were identified in the studied samples, including eight detected for the first time. Nine chemotypes have been identified in P. ernstiae. Parmelia saxatilis shows seven chemotypes, often with galbinic acid and always without fatty acids. Parmelia serrana shows eight chemotypes. Phylogenetic relationships between specimens containing galbinic acid revealed that they do not cluster together. Based on the study new species P. tobolewskiana is proposed.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-47869-w. Introduction Many lichen species produce secondary metabolites, the biological and ecological roles of which have been discussed by various authors1,2. Furthermore, they play an important role in the taxonomy of lichens, particularly in crustose and sterile species, which often have less informative morphological diagnostic characters3–8. Lichen substances are produced by the mycobiont9 and usually deposited as small crystals in the cortex and/or medulla layer on the outer surface of the fungal hyphae2. To date, more than one thousand of these substances have been identified10–12, and thin-layer chromatography (TLC), or alternatively high-performance liquid chromatography (HPLC), has been used to identify them13,14. In the case of the latter, it is more sensitive than TLC and allows the detection of substances present in trace concentrations12,15, which is important for the correct detection of chemosyndromes16. Moreover, recent studies have employed alternative technologies in the chemotaxonomy, such as mass spectrometry (GC-MS, LC-MS, MALDI-MS)17–20. Nevertheless, RP-HPLC (Reverse Phase High-Performance Liquid Chromatography) remains a highly universal method that offers more straightforward and reliable quantitative data, which is particularly advantageous for comparative chemotaxonomic analyses. The genus Parmelia Ach. comprises foliose lichen species with linear, white pseudocyphellae on the upper surface and black rhizines on the lower surface, with or without vegetative propagules21. Up till now, the genus encompasses forty-one species22,23. Nevertheless, it seems probable that the real number is higher, and new species within Parmelia still require formal description24. To date, sixteen lichen metabolites have been identified in Parmelia, with atranorin being the major substance present in most taxa21,27,28. Additionally, the salazinic acid complex, lobaric acid, and a variety of fatty acids have been frequently reported27,29. Less commonly observed are alectorialic and usnic acids, which have only been noted in P. mayi Divakar et al., and P. fraudans (Nyl.) Nyl.21,27,30 However, the number of metabolites in Parmelia taxa varies, with higher number detected in species with true isidia, e.g. P. ernstiae Feuerer & A. Thell, or P. saxatilis (L.) Ach., as well as in a group of taxa without vegetative propagules, such as P. discordans Nyl. and P. pinnatifida Kurok.27,29,31 In contrast, all Parmelia taxa with soredia, e.g., P. sulcata Taylor and P. barrenoae Divakar et al.21,32, and soredioid isidia (i.e. soredia similar to isidia), e.g., P. submontana Nádv. ex Hale and P. sulymae Goward et al., produce only atranorin and the salazinic acid complex27,33,34, and as there are no differences in the chemistry of these species, it is irrelevant as a diagnostic trait29. Moreover, metabolite profiles may vary within a single species. Concerning isidiate Parmelia, certain metabolites have either not been observed in species believed to produce them consistently, or have been detected in taxa for which they have not previously been reported (e.g., lobaric acid)24,25,29,35,36. Despite reports of discrepancies in the chemical composition of isidiate Parmelia, there has been little comprehensive research synthesising the available knowledge on the chemistry of these lichens and their potential chemical variations. The genus Parmelia has been the subject of our research for over a decade29,31,37–44. During our chemical studies, we noticed the presence of orange spots on thin-layer chromatography plates in some isidiate samples. A review of the literature data revealed that these spots may correspond to the presence of galbinic acid (α-acetylsalazinic acid; C20H14O11). This metabolite belongs to β-orcinol depsidones12,45 and has previously been isolated from Hypotrachyna caraccensis (Taylor) Hale, H. enderythraea (Zahlbr.) Hale, Usnea undulata Stirton and U. dasaea Stirton45,46–51. In Parmelia, it has been reported from P. discordans, P. omphalodes (L.) Ach. and P. pinnatifida, but only in trace amounts27,52, and therefore might be barely seen on TLC plates31. Additionally, it has been recently reported from P. rojoi A. Crespo et al.23, as well as from Scottish specimens of P. ernstiae and P. saxatilis collected by Corsie et al.24. However, the latter authors did not provide details of their discovery, despite the fact that this is the first record of this substance in the isidiate species of Parmelia. We therefore decided to carry out further studies with the following goals: (1) to assess the presence of galbinic acid in Parmelia taxa using the RP-HPLC method; (2) to verify if samples containing galbinic acid form a clade within the phylogenetic tree; and (3) to provide the first comprehensive chemical profiles of secondary metabolites present in isidiate Parmelia species. Results and discussion Chemical variability and chemotypes Galbinic acid was detected in Parmelia ernstiae (18 of the 122 examined specimens), P. saxatilis (65 of the 307 examined specimens), and P. serrana A. Crespo et al. (6 of the 168 examined specimens), in the case of the latter species this substance is reported for the first time (Table 1). The studied acid was also identified in the specimens of P. tobolewskiana, a new species described here. In our study, galbinic acid was located between the biosynthetically related substances salazinic and lobaric acids in solvent A, as described by Orange et al.14 and Elix12. In daylight galbinic acid was visible as an orange-brown spot after reaction with sulphuric acid. Table 1Presence/absence (1/0) of secondary metabolites in P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana specimens tested by us, % of samples containing particular metabolite was calculated for each species (see also Table S1). The numbers in the rows indicate the number of specimens producing the metabolite.AtranorinSalazinic acid complexGalbinic acidProtocetraric acidLobaric acidFatty acids01%01%01%01%01%01% Parmelia ernstiae 012210001221001041814.811397.4711594.31810485.2 Parmelia saxatilis 030710003071002426521.2277309.84526285.330700 Parmelia serrana 0168100016810016263.616531.8154148.3716195.8 Parmelia tobolewskiana 02100021001150115002100200 Table 1Presence/absence (1/0) of secondary metabolites in P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana specimens tested by us, % of samples containing particular metabolite was calculated for each species (see also Table S1). The numbers in the rows indicate the number of specimens producing the metabolite.AtranorinSalazinic acid complexGalbinic acidProtocetraric acidLobaric acidFatty acids01%01%01%01%01%01% Parmelia ernstiae 012210001221001041814.811397.4711594.31810485.2 Parmelia saxatilis 030710003071002426521.2277309.84526285.330700 Parmelia serrana 0168100016810016263.616531.8154148.3716195.8 Parmelia tobolewskiana 02100021001150115002100200 Presence/absence (1/0) of secondary metabolites in P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana specimens tested by us, % of samples containing particular metabolite was calculated for each species (see also Table S1). The numbers in the rows indicate the number of specimens producing the metabolite. The obtained chromatograms for the standards of galbinic and norstictic acids showed that under specific analysis conditions, the retention time (Rt) of galbinic acid is 33 min, and for norstictic acid – 34 min (Figs. 1A and 2A). The structure of galbinic and norstictic acids are presented in Figs. 1B and 2B. The absorption spectrum of galbinic acid is very similar to the absorption spectrum of norstictic acid (both have a long-wavelength absorption maximum at about 310 nm) (Figs. 1C and 2C), and is consistent with the previous studies53. The presence of signals with Rt about 33 min and 34 min on the chromatograms of the ten studied samples would suggest the presence of galbinic or norstictic acids (Fig. 3). To rule out the possibility of the coincidental match of retention times, the absorption spectra of the analyzed signals from the chromatogram were also compared. The presence of a signal around 33 min (Fig. 3), whose UV spectrum matched that of the reference sample (Fig. 1), indicates the presence of galbinic acid in the: P. saxatilis samples: UGDA L-24348, UGDA L-24349, UGDA L-24350, UGDA L-24276, UGDA L-49666, and UGDA L-49675 (Fig. 3). In the samples of P. serrana (UGDA L-49661 (Fig. 3), UGDA L-49659 and UGDA L-53822) and P. ernstiae (UGDA L-61370) galbinic acid was not found. Moreover, the obtained results indicate that galbinic acid definitely dominates quantitatively in two samples, UGDA L-24348 and UGDA L-24349, and clearly in one sample UGDA L-24350 (Fig. 3). On the other hand, the lack of a signal Rt = 34 min on the chromatograms of all the extracts means that these samples do not contain norstictic acid (Fig. 3). Fig. 1Chromatogram of galbinic acid standard (Rt=33.608 min) recorded at λ = 254 nm (A) with structure (B) and absorption spectrum (with long-wavelength absorption maximum at λmax 310 nm) (C). Chromatogram of galbinic acid standard (Rt=33.608 min) recorded at λ = 254 nm (A) with structure (B) and absorption spectrum (with long-wavelength absorption maximum at λmax 310 nm) (C). Fig. 2Chromatogram of norstictic acid standard (Rt=34.369 min) recorded at λ = 254 nm (A) with structure (B) and absorption spectrum (with long-wavelength absorption maximum at λmax 310 nm) (C). Chromatogram of norstictic acid standard (Rt=34.369 min) recorded at λ = 254 nm (A) with structure (B) and absorption spectrum (with long-wavelength absorption maximum at λmax 310 nm) (C). Fig. 3Chromatograms for P. saxatilis (UGDA L-24348, UGDA L-24349, UGDA L-24350, UGDA L-24276, UGDA L-49666, UGDA L-49675), and P. serrana (UGDA L-49661) specimens recorded at λ = 254 nm (retention time of galbinic acid is marked with red frames). Chromatograms for P. saxatilis (UGDA L-24348, UGDA L-24349, UGDA L-24350, UGDA L-24276, UGDA L-49666, UGDA L-49675), and P. serrana (UGDA L-49661) specimens recorded at λ = 254 nm (retention time of galbinic acid is marked with red frames). Galbinic acid is a substance rarely observed in isidiate Parmelia (Table 2). Indeed, only Corsie et al.24 found it in P. saxatilis and P. ernstiae specimens. Interestingly, in their study, more specimens of P. saxatilis were reported with galbinic acid than with lobaric acid24. Our results show the opposite proportion (Table 1). Table 2Selected previous and present overview of chemical characteristics of species within the Parmelia saxatilis group (ATR – atranorin, SAL – salazinic acid with consalazinic acid, LOB – lobaric acid, GAL – galbinic acid, FAT – fatty acids, PROTO – protocetraric acid ). New chemotypes are in bold.TaxaChemotype numberChemotypeReferences Parmelia ernstiae I ATR, SAL This study IIATR, SAL, LOBMolina et al. (2004); Corsie et al. (2019); This studyVATR, SAL, LOB, FATThell et al. (2008); Molina et al. (2011); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This study VI ATR, SAL, FAT This study VIIATR, SAL, LOB, GAL, FATCorsie et al. (2019); This study IX ATR, SAL, LOB, PROTO This study X ATR, SAL, LOB, GAL, PROTO This study XIIATR, SAL, LOB, FAT, PROTOThell et al. (2008); Molina et al. (2011); Corsie et al. (2019); This study XIV ATR, SAL, LOB, GAL, FAT, PROTO This study Parmelia saxatilis IATR, SALMolina et al. (2004); Ossowska et al. (2018); Crespo et al. (2020); This studyIIATR, SAL, LOBMolina et al. (2004); Thell et al. (2008); Ossowska et al. (2018); Tsurykau et al. (2019); Corsie et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This studyIIIATR, SAL, LOB, GALCorsie et al. (2019); This studyIVATR, SAL, GALCorsie et al. (2019); This studyIXATR, SAL, LOB, PROTOThell et al. (2008); Molina et al. (2011); Corsie et al. (2019); This studyXATR, SAL, LOB, GAL, PROTOCorsie et al. (2019); This studyXIATR, SAL, GAL, PROTOCorsie et al. (2019); This study Parmelia serrana IATR, SALMolina et al. (2004); Ossowska et al. (2018); Corsie et al. (2019); This studyIIATR, SAL, LOBMolina et al. (2004); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); This studyVATR, SAL, LOB, FATOssowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); This studyVIATR, SAL, FATThell et al. (2008); Molina et al. (2011); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This study VII ATR, SAL, LOB, GAL, FAT This study VIII ATR, SAL, GAL, FAT This study IX ATR, SAL, LOB, PROTO This study XIIIATR, SAL, FAT, PROTOThell et al. (2008); Molina et al. (2011) Parmelia tobolewskiana II ATR, SAL, LOB This study X ATR, SAL, LOB, GAL, PROTO This study * Thell et al. (2008) and Corsie et al. (2019) checked specimens for galbinic and protocetraric acids, however they did not find them in particular samples. Table 2Selected previous and present overview of chemical characteristics of species within the Parmelia saxatilis group (ATR – atranorin, SAL – salazinic acid with consalazinic acid, LOB – lobaric acid, GAL – galbinic acid, FAT – fatty acids, PROTO – protocetraric acid ). New chemotypes are in bold.TaxaChemotype numberChemotypeReferences Parmelia ernstiae I ATR, SAL This study IIATR, SAL, LOBMolina et al. (2004); Corsie et al. (2019); This studyVATR, SAL, LOB, FATThell et al. (2008); Molina et al. (2011); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This study VI ATR, SAL, FAT This study VIIATR, SAL, LOB, GAL, FATCorsie et al. (2019); This study IX ATR, SAL, LOB, PROTO This study X ATR, SAL, LOB, GAL, PROTO This study XIIATR, SAL, LOB, FAT, PROTOThell et al. (2008); Molina et al. (2011); Corsie et al. (2019); This study XIV ATR, SAL, LOB, GAL, FAT, PROTO This study Parmelia saxatilis IATR, SALMolina et al. (2004); Ossowska et al. (2018); Crespo et al. (2020); This studyIIATR, SAL, LOBMolina et al. (2004); Thell et al. (2008); Ossowska et al. (2018); Tsurykau et al. (2019); Corsie et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This studyIIIATR, SAL, LOB, GALCorsie et al. (2019); This studyIVATR, SAL, GALCorsie et al. (2019); This studyIXATR, SAL, LOB, PROTOThell et al. (2008); Molina et al. (2011); Corsie et al. (2019); This studyXATR, SAL, LOB, GAL, PROTOCorsie et al. (2019); This studyXIATR, SAL, GAL, PROTOCorsie et al. (2019); This study Parmelia serrana IATR, SALMolina et al. (2004); Ossowska et al. (2018); Corsie et al. (2019); This studyIIATR, SAL, LOBMolina et al. (2004); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); This studyVATR, SAL, LOB, FATOssowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); This studyVIATR, SAL, FATThell et al. (2008); Molina et al. (2011); Ossowska et al. (2018); Corsie et al. (2019); Tsurykau et al. (2019); Crespo et al. (2020); Castellani et al. (2021); This study VII ATR, SAL, LOB, GAL, FAT This study VIII ATR, SAL, GAL, FAT This study IX ATR, SAL, LOB, PROTO This study XIIIATR, SAL, FAT, PROTOThell et al. (2008); Molina et al. (2011) Parmelia tobolewskiana II ATR, SAL, LOB This study X ATR, SAL, LOB, GAL, PROTO This study * Thell et al. (2008) and Corsie et al. (2019) checked specimens for galbinic and protocetraric acids, however they did not find them in particular samples. Selected previous and present overview of chemical characteristics of species within the Parmelia saxatilis group (ATR – atranorin, SAL – salazinic acid with consalazinic acid, LOB – lobaric acid, GAL – galbinic acid, FAT – fatty acids, PROTO – protocetraric acid ). New chemotypes are in bold. * Thell et al. (2008) and Corsie et al. (2019) checked specimens for galbinic and protocetraric acids, however they did not find them in particular samples. Substances always found in P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana were atranorin and the salazinic acid complex (salazinic and consalazinic acids). Additionally, lobaric acid (present in 262 samples of P. saxatilis, 115 of P. ernstiae, 14 of P. serrana and 2 of P. tobolewskiana) was also observed (Table 1). Parmelia ernstiae and P. serrana often produced fatty acids, which were always absent in P. saxatilis and P. tobolewskiana samples (Table 1, Supplementary Material Table S1). In this study, we also identified the presence of protocetraric acid in all taxa (Table 1). In contrast, protocetraric acid was only identified in specimens from the Scotland in P. saxatilis and in one specimen of P. ernstiae24. Corsie et al.24 reported fumarprotocetraric acid (a derivative of protocetraric acid), which we did not find in our samples. Concerning protocetraric acid, our results, together with data from the literature, confirm that it is present in Parmelia spp. in trace amounts, hence it may not be detectable on TLC plates24,25,36. According to the literature data24,27, trace amounts of other metabolites, such as chloroatranorin, nephrosterinic or isonephrosterinic acids, may also be present in species treated here. However, it is important to note that the presence of these additional substances is often omitted in various works23,25,29,30,36,54. Chloroatranorin is a derivative of atranorin, with a similar Rf class. According to Elix12, it is only visible on a TLC plate if it is present in the sample at a high concentration. Nephrosterinic acid and its derivative isonephrosterinic acid have been found in P. ernstiae and P. serrana, but only in minor amounts and using RP-HPLC27. Selected previous and current chemical characteristics of analyzed species are summarized in Table 2. In this table, we have highlighted the new chemotypes of P. ernstiae and P. serrana in bold. Based on TLC and RP-HPLC results, we defined 14 chemotypes in the four analyzed species (Fig. 4). One chemotype: atranorin, the salazinic acid complex and protocetraric acid was not observed in any of the examined specimens, but noted in P. saxatilis by Corsie et al.24. All four species were similarly diverse chemically, and P. ernstiae is characterized by nine chemotypes (among them chemotype V is the most common), P. saxatilis by seven chemotypes (among them chemotype II is predominant), P. serrana by eight chemotypes and P. tobolewskiana by two chemotypes (but only two specimens were analyzed) (for detailed description of the chemotypes see Fig. 4; Table 2). According to Feuerer and Thell55, the chemotype V is diagnostic for P. ernstiae. However, in some specimens, the absence of lobaric acid and/or fatty acids, and/or the presence of galbinic acid was observed. In P. ernstiae, chemotypes I, VI, IX, X and XIV are observed for the first time. The chemical composition of P. saxatilis observed in our study agrees with previous reports and seems characteristic of this species21,27,33,54. However, we noted the absence of lobaric acid in some samples and the presence of galbinic acid instead, as well as the absence of both substances in some specimens. In case of P. serrana we found atranorin, the salazinic acid complex and fatty acids in most specimens, which is consistent with Molina et al.54. We also observed some variation (Fig. 4), e.g. fatty acids were absent in seven specimens, one of which additionally contained lobaric acid. Chemotypes VII, VIII and IX are new to P. serrana (Fig. 5). Fig. 4Collapsed phylogenetic tree of Parmelia saxatilis group, with the clade of the new species P. tobolewskiana marked in green box (the specimen P. cf. tobolewskiana was no included in the chemical analyses). On the right side of the phylogenetic tree pie charts indicate the percentage of certain chemotypes of P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana specimens (tree top view) observed in: A – specimens used for molecular testing, B – all samples analyzed by us during our study, but not confirmed by phylogenetic evidence (see also Fig. S1). Chemotypes I, VI, IX, X and XIV are new to P. ernstiae and VII, VIII and IX in P. serrana. Collapsed phylogenetic tree of Parmelia saxatilis group, with the clade of the new species P. tobolewskiana marked in green box (the specimen P. cf. tobolewskiana was no included in the chemical analyses). On the right side of the phylogenetic tree pie charts indicate the percentage of certain chemotypes of P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana specimens (tree top view) observed in: A – specimens used for molecular testing, B – all samples analyzed by us during our study, but not confirmed by phylogenetic evidence (see also Fig. S1). Chemotypes I, VI, IX, X and XIV are new to P. ernstiae and VII, VIII and IX in P. serrana. Fig. 5Haplotype networks showing relationships between nuITS rDNA sequences from (A) Parmelia ernstiae; (B) P. saxatilis; and (C) P. serrana. The colors on the network correspond to the certain chemotypes described in Fig. 3, while the numbers 1, 2, 3 and 4 correspond to the haplotype number. Mutational changes are shown as hatch marks on the lines between haplotypes. Haplotype networks showing relationships between nuITS rDNA sequences from (A) Parmelia ernstiae; (B) P. saxatilis; and (C) P. serrana. The colors on the network correspond to the certain chemotypes described in Fig. 3, while the numbers 1, 2, 3 and 4 correspond to the haplotype number. Mutational changes are shown as hatch marks on the lines between haplotypes. To test phylogenetic relationship between different chemotypes, we constructed haplotype networks (Fig. 4). For these, we used all our P. ernstiae, P. saxatilis and P. serrana sequences, in which chemistry had been checked by us (Table S1 and S2). Parmelia ernstiae sequences group into two haplotypes that differ by two nucleotide substitutions. The first haplotype contains nine specimens with chemotypes I, II, V and VII, while the second contains twenty-seven specimens (chemotypes I, II, V, VI, VII, X and XIV). In P. saxatilis, four haplotypes were found to differ by one or two nucleotide substitutions. The majority of the samples belong to one haplotype (forty-four samples), representing seven chemotypes. The specimens with galbinic acid (chemotypes III, X, XI, and XII) are represented by three haplotypes. Parmelia serrana has two haplotypes that differ by one substitution. Chemotype VI is represented by two haplotypes, while the other chemotypes have one haplotype (Fig. 4; Table 3). For the MCM7 marker, the analysis of the haplotypes showed no differences between the samples of the analyzed taxa. In summary, we did not observe any pattern in which individual chemotypes were associated with a particular haplotype, or in which the number of differences between them was significant. Similar observations have been described by Guzow-Krzemińska et al.6 in Lepra amaroides (H. Magn.) I. Schmitt, Hodkinson & Lumbsch samples differing in the presence/absence of hypothamnolic or barbatic acid. Table 3Haplotype network results showing the number of samples per chemotype grouped into observed haplotypes. Description of each chemotype is in Fig. 4.Parmelia ernstiaeParmelia saxatilisParmelia serranaNumber of haplotypes12123412Number of samplesChemotypesI12 1 61II3719 2III19IV4V414 6VI1 1 25VII11 1IX1 31X1 2XI 1XIV4 Table 3Haplotype network results showing the number of samples per chemotype grouped into observed haplotypes. Description of each chemotype is in Fig. 4.Parmelia ernstiaeParmelia saxatilisParmelia serranaNumber of haplotypes12123412Number of samplesChemotypesI12 1 61II3719 2III19IV4V414 6VI1 1 25VII11 1IX1 31X1 2XI 1XIV4 Haplotype network results showing the number of samples per chemotype grouped into observed haplotypes. Description of each chemotype is in Fig. 4. The phylogenetic tree shows all the studied Parmelia species arranged into eleven clades (Fig. 4 and Fig. S1): six with species with true isidia: P. saxatilis, P. rojoi, lineage named here as P. tobolewskiana, P. ernstiae, P. serrana and P. mayi; two with soredioid isidia: P. sulymae and P. submontana intermingled with P. hygrophila Goward & Ahti (separation of these species based on ITS sequences is not supported; see also Crespo et al.23); three with species that do not produce vegetative propagules: P. imbricaria Goward et al., P. pinnatifida and P. omphalodes intermingled with P. discordans (separation of these species based on ITS sequences is not supported; see also Ossowska et al.31). The chemistry of each specimen of P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana analyzed here was indicated (Fig. 4 and Fig. S1). Hidden species diversity in Parmelia Cryptic species as defined by Bickford et al.56 and Crespo and Pérez-Ortega57, are species that were previously classified as a single, nominal taxon due to their superficial morphological and/or chemical similarities, but molecular evidence indicated that they are genetically distinct. Examples of this phenomenon are the recently described taxa Parmelina clandestina Barcenas-Peña et al. and P. mediterranea Barcenas-Peña et al. separated from P. tiliacea (Hoffm.) Hale58. However, there are many more such cryptic species, and the widespread use of genetic data in lichenology has contributed significantly to their identification7,59,60. Recent work shows that comprehensive analyses, including statistical test, have revealed the existence of small, subtle or previously unnoticed differences in many cryptic species. In many cases, these differences are not significant in the context of field research or taxonomic keys and are only apparent to experts61. Such examples include the shape of the basidiomes in Sulzbacheromyces chocoensis Coca et al.62 or the size of the thallus in Cora garagoa Simijaca et al.63. Therefore, when such subtle differences are observed the more accurate term is near or semi-cryptic species61. According to the definition proposed by Dal Forno et al.64, near-cryptic species (or kapo- (phylo-) cryptic) have a very low phenotypic distance, which in practice means that the species are indistinguishable on the basis of phenotypic characteristics. In contrast, semi-cryptic (or allo- (phylo-) cryptic) exhibit a greater phenotypic distance, signifying that the taxa are highly similar to each other, yet can be distinguished through meticulous analysis64. It is important to note that this category also encompasses differences in geographical distribution and ecological requirements61. For instance, Xanthoparmelia tasmanica (Hook. f. & Taylor) Hale from Australia and New Zealand, is a semi-cryptic species to X. hypofusca (Gyeln.) B.P. Hodk. & Lendemer from North America66, and Caloplaca nigromarina Vondrák, Říha, Arup & Søchting which occur only on siliceous rocks on coastal cliffs, is semi-cryptic to C. flavocitrina (Nyl.) H. Olivier occurring on tree bark, wood, and mostly calcareous rocks65. In Parmelia, a well-documented example of a cryptic species is P. encryptata A. Crespo, Divakar & M.C. Molina26,39,44,67. Contrary to this, the most recent studies4068 suggest that there are minor discrepancies in the arrangement of rhizines and the pigmentation of the lower thallus in P. encryptata in comparison to P. sulcata. Consequently, the debate about whether P. encryptata should be classified as a cryptic species remains unresolved. In the case of the isidiate Parmelia species discussed here, we believe that they are semi-cryptic due to their overlapping morphology and minor chemical differences compared to P. saxatilis23,24,29,34,40,54. Parmelia tobolewskiana has the same morphology and chemical profiles as P. saxatilis; however, we observed subtle differences between these two species. In P. saxatilis isidia are abundant and distributed irregularly on the upper surface, with lobes reaching up to 4 mm in width. In the new species, isidia are mainly marginal, with only a few laminal arranged in strips (or lines) along the pseudocyphellae and lobes are narrower (up to 2 mm). The two species exhibit also similar ecological requirements and probably geographical distribution. Parmelia saxatilis is a widely distributed species21,6769, common in Poland29,31, occurring on a variety of substrate type, and in some regions, such as the Nordic countries, it prefers rock substrates33. Parmelia tobolewskiana has so far only been collected on stones and only in Poland and Spain. However, its known distribution may change as further studies are conducted. Because the few features which we recognized are very subtle and probably can not be detected by lichenologists who do not specialize in Parmelia, we can conclude that the new species can only be distinguished from P. saxatilis by analyzing the molecular data. Parmelia ernstiae and P. serrana are examples of semi-cryptic species. These species were differentiated from P. saxatilis based on molecular evidence, with initial literature data defining the morphological and chemical features that distinguish them27,30,33,54,55. However, a larger sampling demonstrated that they exhibit high intraspecies variability and low interspecies variability24,25,29,35,40. Furthermore, many of the traits previously identified as diagnostic are in fact non-diagnostic. This was exemplified by the presence of a pruina on the upper surface. Strongly pruinose thallus was identified as a key feature of P. ernstiae55, while P. saxatilis and P. serrana were characterized as having partially pruinose to epruinose thalli54. Phylogenetic analysis demonstrated that the pruina is a feature of a limited value in the identification of Parmelia species29. Similar considerations, but based on a different set of features, such as the shape and distribution of isidia, the presence of lobules, or the color and shape of the tips, were documented by Corsie et al.24. Concurrently, minor morphological disparities between P. ernstiae, P. saxatilis and P. serrana can be identified. For example, the lobe tips are very rarely brown at edge and fading proximally in P. saxatilis, and P. serrana is characterized by the absence of lobules24. Concerning chemistry, the composition of secondary metabolites has been previously identified as important feature to distinguish Parmelia species21,27,33,54,55. One of the earliest hypotheses claimed that P. serrana does not contain lobaric acid and this distinguished it from P. saxatilis and P. ernstiae54(see also Table 2). However, it was demonstrated that it may be present in some specimens of P. serrana24,25,29,71 and the absence of this metabolite was observed in some specimens of P. ernstiae and P. saxatilis24,29. Additionally, analogous observations regarding the absence of fatty acids have been documented for P. ernstiae and P. serrana29. All these chemotypes were also recorded in this study (Fig. 4; Table 1). It is worth mentioning that we did not find any fatty acids in all P. saxatilis and P. tobolewskiana samples analyzed here. However, Corsie et al.24 reported the presence of these substances in one specimen of P. saxatilis. Based on these data (307 samples analyzed here vs. 78 in Corsie et al.24) we suppose that the result obtained for this sample could be due to contamination during TLC, in which case it would require re-analysis. The lack of fatty acids in P. saxatilis seems to be one of the most important differences between this species and P. ernstiae and P. serrana, as well as P. mayi and P. rojoi, but not P. tobolewskiana. A similar problem is observed with the absence of atranorin in several samples analyzed by Corsie et al.24. This is likely attributable to insufficient concentration; however, it is advisable to reiterate these samples. The value of chemotaxonomy in lichen identification has been well documented since Nylander12,72–74, and according to Hawksworth75, any taxonomic revision without chemical information is incomplete. In the past, differences in secondary metabolites have been frequently employed as a criterion for the identification of novel species or varieties76. However, with the advent of molecular techniques that provide an additional set of traits used in lichen identification, the validity of such findings has been reassessed. For example, in Lepra subventosa (Malme) I. Schmitt & Lumbsch, three varieties were previously distinguished77, one of which is now classified as a species: L. pseudosubventosa Kukwa & B. Guzow-Krzem.6 Similarly, Parmeliopsis ambigua (Hoffm.) Nyl. and P. hyperopta (Ach.) Vain., initially proposed as a chemical varieties, are distinct species78. In contrast, genetic data helped to merge chemical varieties in Ramalina siliquosa complex, e.g., R. stenoclada W.L. Culb. or R. atlantica W.L. Culb., described by Culberson and Culberson79, into a single species, R. cuspidataNyl.80 Here, we have shown a huge intraspecific chemical variation in P. ernstiae, P. saxatilis, P. tobolewskiana or P. serrana. Taxonomy Parmelia tobolewskiana Ossowska, Kukwa, Szczepańska, Szymczyk & Guzow-Krzem., sp. nov. MycoBank No: MB860037. Figure 6. Fig. 6Morphology of Parmelia tobolewskiana. A–B sublinear lobes with marginal and laminal pseudocyphellae; C simple rhizines; D. marginal and laminal isidia together with lobules. Scale bars: 1 mm. Morphology of Parmelia tobolewskiana. A–B sublinear lobes with marginal and laminal pseudocyphellae; C simple rhizines; D. marginal and laminal isidia together with lobules. Scale bars: 1 mm. Diagnosis Semi-cryptic species differing from P. saxatilis in its phylogenetic position, and characterized by narrow (up to 2 mm in width), sublinear lobes, isidia developing mainly on margins, and the production of atranorin, the salazinic acid complex, together with lobaric, protocetraric and galbinic acids. Type Poland. Garb Gielniowski; Skałki Piekło pod Niekłaniem nature reserve, 51.18400°N, 20.66680°E, rocky outcrop in oak-fir forest (sandstone), saxicolous, 31 May 2023, R. Szymczyk s.n. (holotype UGDA L-65664 = ITS: PV921432, MCM7: PV946988; chemotype II). Description Thallus foliose, saxicolous, loosely to moderately attached, up to 8 cm. Lobes sublinear, branched, narrow, to 2 mm wide, contiguous. Upper surface grey to grey-brown, with brown tips, smooth to shallowly pitted, epruinose. Pseudocyphellae marginal and laminal, linear to irregular, not connected or very few reticulate, very often serving as a place of origin for isidia. Isidia numerous, mainly marginal and laminal arranged in lines, globulose to cylindrical, simple to branched, darkening at the apices, up to 0.13 mm diam, up to 0.5 mm long; marginal lobules present, sparse. Soredia, pustules and cilia absent. Medulla white. Lower surface black to brown at margins; rhizines up to 1.5 mm long, black, simple to furcate. Apothecia absent. Pycnidia laminal, immersed, up to 140 μm in diameter, with a tendency to group toward the lobe margins; pycnidial wall dark grey, K+ brown; conidia 6–7 × 1 μm. Secondary chemistry Atranorin, salazinic and consalazinic acids, lobaric acid, often with protocetraric acid and galbinic acid. Habitat and distribution Parmelia tobolewskiana is a saxicolous lichen. The species was found in two localities in the mountain and upland regions of Poland, and one in Spain26. However, we did not see a specimen from Spain and used only the sequence present in GenBank. Etymology Named in honor of the lichenologist, Zygmunt Tobolewski, for his enormous contribution to the knowledge of lichens in Poland. Additional material examined Poland. Western Sudety; Rudawy Janowickie, Krzyżna Góra, saxicolous, 25 Sept. 2022, K. Szczepańska s.n. (UGDA L-61392 = ITS: PV921431, MCM7: PV946987; chemotype X). Notes Parmelia tobolewskiana is a semi-cryptic species closely related to P. saxatilis (Fig. 4) and P. rojoi. In the phylogenetic tree, it forms a well-supported monophyletic clade related to P. saxatilis and P. rojoi, but without support (Fig. 4 and Fig. S1). Besides two specimens from Poland, this clade also includes one specimen from Spain, previously deposited by Moya et al.26. Although we did not see a specimen from Spain, we decided to describe this lineage as a novel taxon named here as P. tobolewskiana. We identified at least ten autapomorphic nucleotides in the nuITS rDNA region that differentiate P. tobolewskiana from P. rojoi and P. saxatilis; the study regarding the haplotype differences between these species and other isidiate Parmelia will be treated in the forthcoming paper. We did not observe any differences in the MCM7 sequences of these species. The new species differs from P. saxatilis in the distribution of isidia, which are mainly marginal, with only a few laminal arranged in strips along the pseudocyphellae, furthermore, the lobes are narrower (up to 2 in width, and up to 4 mm in P. saxatilis). In contrast, in P. rojoi, the lobes are broader, up to 4.5 mm, the isidia are mainly laminal and lobaric acid was not observed23. All other isidiate Parmelia species, with the exception of P. saxatilis and P. rojoi, are also morphologically distinct from P. tobolewskiana. In detail, the lobes of the new species are narrow, up to 2 mm wide, and up to 3 mm in P. serrana; up to 4 mm in P. mayi, and up to 8 mm in P. ernstiae29,30. Furthermore, in P. tobolewskiana the isidia are mainly marginal, with few laminal developed from pseudocyphellae. In contrast, in P. ernstiae and P. mayi the isidia are mainly laminal29,30,55 and laminal and marginal in P. serrana24,54. In addition, P. mayi produces also alectorialic acid34. The new taxon differs from other species of the P. saxatilis clade defined by Molina et al.34 in the type of vegetative propagules. Parmelia submontana, P. hygrophila and P. sulymae have soredioid isidia21,34. On the other hand, P. omphalodes, P. discordans, P. pinnatifida, and P. imbricaria do not produce vegetative propagules at all31,34. Another isidate Parmelia species is P. squarrosa Hale, which can be confused with P. tobolewskiana due to the presence of narrow lobes with marginal isidia. However, P. squarrosa has squarrose rhizines and therefore probably belongs to the P. sulcata clade (all species in the P. saxatilis group have simple to branched rhizines). In addition, P. tobolewskiana can also develop laminal isidia and produces lobaric and galbinic acids, which are always absent in P. squarrosa. The latter species is corticolous, known from North America and Japan21,54. Recently, a new isidiate species, Parmelia neosaxatilis Ossowska, Kosecka & Kukwa, has been described. Although it is morphologically and chemically similar to P. saxatilis, its phylogeny indicates a close relationship with P. aff. omphalodes81. A distinguishing trait that separates this species from P. tobolewskiana is the structure of the pseudocyphellae, which are connected and form a white reticulate network. In contrast, in P. tobolewskiana, the majority of pseudocyphellae are not connected, with only a few reticulate. In addition, in P. neosaxatilis, the isidia are laminal and predominantly located in the central part of the thallus, whereas in P. tobolewskiana, isidia are laminal and marginal, and arranged in line. Parmelia neosaxatilis is presumably endemic to Sardinia81. Materials and methods Taxon sampling In the course of our studies, we analyzed chemically all the isidiate Parmelia species that we had investigated over ten years. Some of them were too old for molecular studies. In these cases, we used morphological, chemical and ecological approaches to identify species, as recommended by Corsie et al.24. In total, 599 lichen specimens of the species P. ernstiae, P. saxatilis, P. serrana and P. tobolewskiana deposited in B, GPN, KRAM, KRAP, KTC, LBL, LD, LOD, MSC, POZ, S-F, SLTC, UGDA, UPS, WA, and WRSL herbaria were used for chemical analyses employing the TLC method in solvents A and C14. A morphological examination was conducted using a stereomicroscope following the traits proposed by Ossowska et al.29 and Corsie et al.24. Detailed localities and chemical results are presented in Supplementary Data Table S1. Chemical methods The reference of galbinic acid was isolated from a specimen of Hypotrachyna microblasta (Vain.) Hale (UGDA L-18205). Solvent A was used and the preparative thin-layer chromatography method was further employed14. Then, the plate was observed under UV 254 nm and the galbinic acid spots were marked with a pencil. A layer of silica gel was subsequently scraped from the plate and extracted in acetone. The obtained solution was used for RP-HPLC as a reference sample. A similar protocol was used for the norstictic acid reference sample, but isolated from Pleurosticta acetabulum (Neck.) Elix & Lumbsch (UGDA L-52120). We decided to use a norstictic acid sample, as a reference, because both substances show similar properties, i.e. the spots on TLC plates are orange and have a similar Rf class12. Based on the TLC results, we selected six samples of P. saxatilis with galbinic acid. Lichen extracts were isolated using preparative TLC in the same way as previously described. Furthermore, we added four additional samples of P. ernstiae and P. serrana in which galbinic acid was not visible on the TLC plate to verify whether galbinic acid may be present in Parmelia spp. in trace amounts. Ten samples of P. ernstiae, P. serrana and P. saxatilis were subjected to qualitative reverse-phase high-performance liquid chromatography (RP-HPLC) analysis using a Shimadzu LC-20AD system. Reverse-phase high-performance liquid chromatography (RP-HPLC) is widely used for chemical profiling of lichen extracts as well as for the identification and quantitative determination of their secondary metabolites82–87. Separation was carried out on a C8 reverse-phase column (250 × 4.6 mm, 5 μm particle size; Supelco Discovery Analytical). Elution was performed using a linear gradient of solvent B (0–100%) over 60 min at a flow rate of 1 ml/min. Solvent A consisted of water containing 0.1% TFA, whereas solvent B was 80% acetonitrile in water with 0.08% TFA. Prior to each injection, the column was equilibrated with solvent A for at least 15 min. Detection was carried out using a photodiode array detector (SPD-M20A, Shimadzu), and spectra were recorded in the range of 190–800 nm during the RP-HPLC run. Samples containing galbinic and norstictic acid standards were analyzed under the same conditions. The presence of these compounds in the analyzed samples was confirmed based on their retention times and corresponding UV–Vis spectra. Chromatographic data were processed using LCsolution software (version 1.23 SP1, Shimadzu). DNA isolation, PCR amplification, and sequencing For molecular studies, we used 42 Parmelia specimens, including 13 specimens in which we identified galbinic acid using TLC. DNA was isolated using Sherlock AX Kit (A&A Biotechnology, Poland) as in Ossowska et al.29. Fungal nuITS rDNA was amplified using the primer sets ITS5 and ITS488 and the MCM7 gene employing X_mcm7_F and X_mcm7_R primers89. The following PCR cycling parameters were applied to amplify nuITS region: initial denaturation at 95 °C for 3 min, followed by 35 cycles at 94 °C for 30 s, 54 °C for 30 s (for ITS5 and ITS4). In the case of MCM7 region the following PCR conditions were applied: initial denaturation at 95 °C for 5 min followed by 38 cycles at 94 °C for 45 s, 56 °C for 50 s and 72 °C for 1 min, and the final extension at 72 °C for 5 min. PCR products were purified using a Clean-Up System Kit (A&A Biotechnology, Poland) and sequenced by Macrogen service (http://www.macrogen-europe.com). New sequences of Parmelia species have been deposited in the GenBank database (Table S2). We also used 95 sequences of P. ernstiae, P. saxatilis and P. serrana published by us previously29,31,40,41 and 34 reference sequences of taxa from the P. saxatilis group23,24,26,30,34,54. The sequences of P. sulcata were used as an outgroup (Table S2). Phylogenetic analyses The newly generated sequences and selected representatives of Parmelia spp. from GenBank were automatically aligned using Seaview90,91 using the algorithm MUSCLE92, followed by manual correction and elimination of terminal ends. Unambiguous characters were eliminated by employing less stringent parameters in Gblock93 using Seaview software91. W-IQ-Tree94 was used to perform phylogenetic analyses of each marker employing ultrafast bootstrap. No conflict between highly supported nodes was found. Then nuITS and MCM7 were concatenated. The final alignment consisted of 139 sequences of 1018 sites. Further, W-IQ-Tree was used for model selection and SYM+G4 for nuITS rDNA marker and K2P+G4 model for MCM7 gene were used for the phylogenetic analysis under the maximum likelihood criterion94–97. Support values were calculated with 100 standard bootstrap replicates96. Bayesian analysis was carried out using a Markov Chain Monte Carlo (MCMC) method, in MrBayes v. 3.2.698,99 on the CIPRES Web Portal100. Two parallel MCMC runs were performed, each using four independent chains and ten million generations, sampling every 1000th tree. Posterior probabilities (PP) were determined by calculating a majority-rule consensus tree after discarding the initial 25% trees of each chain as the burn-in. The convergence of the chains was confirmed by the convergent diagnostic of the Potential Scale Reduction Factor (PSRF), which approached 1 and the ‘average standard deviation of split frequencies’ was < 0.01. Phylogenetic trees were visualized using FigTree v.1.4.2101, selected branches were collapsed and final trees were prepared in Inkscape (https://inkscape.org/). Haplotype network For the haplotype network analysis for each marker, the sequences of Parmelia ernstiae, P. saxatilis, and P. serrana obtained during the present study and published by us in earlier work29,31,40,41 were used (see Tables S1 and S2). All sequences were aligned in Seaview91, and the terminal ends were cut. The final nuITS rDNA alignment consisted of 39 sequences of P. ernstiae with 500 sites, 48 sequences of P. saxatilis with 470 sites, 37 sequences of P. serrana with 484 sites, while MCM7 alignment consisted of 7 sequences of P. ernstiae with 484 sites, 15 sequences of P. saxatilis with 492 sites, and 5 sequences of P. serrana with 493 sites. The TCS network102 was created using PopART software (http://popart.otago.ac.nz) and modified in Inkscape (http://inkscape.org). Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2