We sampled 28 specimens, representing species of the European “Flaccidum” and “Nigrescens”-groups from Scandinavia and including some extra-Scandinavian material for comparison. The collections resulting from our own recent fieldwork are deposited in the herbarium S, with additional material studied from GZU, hb. Malíček, TBS, UPS and S. Collections and the sequences used are summarized in Table 2. Herbarium acronyms follow Thiers (2018).

Anatomical features were studied using a light microscope on thin sections cut with a razor blade or squash preparations mounted in water. Measurements of mature spores were taken outside of the asci under × 1000 magnification using oil immersion with a precision of 0.5 μm, or from calibrated digital photographs using NIS-Elements (Nikon, Japan) with a precision of 0.1 μm. Spore measurements are presented in the format: (minimum value observed–) range including 80% of the observed values (–maximum value observed), with the mean of all observed values in the center and italicized (Table 1). Full lists of specimens examined in this study with DNA voucher codes and GenBank Accession numbers for newly generated sequences are given in Table 2.

Two apothecia with surrounding thalline parts, or a thallus fragment in the case of sterile samples, were selected for extraction. We extracted total DNA using the Plant DNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions except in order to increase the concentration of DNA, we used half the amount of Elution buffer in the last step. We amplified one mitochondrial ribosomal and three nuclear protein-coding genes. Approximately 0.8 kb of the small subunit of the mitochondrial rDNA (mtSSU) was amplified using the primers mrSSU1 and mrSSU3R (Zoller et al. 1999). Approximately 0.6 kb of Beta-tubulin (b-tub) was amplified and sequenced using the primers Bt3-LM5 and Bt10-LM3 (Myllys et al. 2001) and BetaCollF and BetaCollR (Otálora et al. 2013a). The PCR primers were used in the following combinations: Bt3-LM5 and Bt10-LM3, BetaCollF and BetaCollR, Bt3-LM5 and BetaCollR, BetaCollF and Bt10-LM3 (the best working combination), BetaCollF and BetaColl-intR, and BetaColl-intF2 and Bt10LM3. About 0.6 kb of mini–chromosome maintenance complex component 7 DNA replication licensing factor (MCM7) was amplified and sequenced using the primers Mcm7-709for and Mcm7-1348rev (Schmitt et al. 2009). The locus RNA polymerase II protein coding gene (RPB2, ~2 kb) was amplified as two parts: ~ 1 kb for RPB2 5–7 and ~ 1 kb for RPB2 7–11. Amplification was performed using the primers fRPB2-5F, fRPB2-7cF, fRPB2-7cR and fRPB2-11aR (Liu et al. 1999), which were also used as sequencing primers. PCR amplifications were performed using Illustra™ Hot Start PCR beads, according to manufacturer’s instructions. PCR reactions were performed using the same settings as in previous studies (Košuthová et al. 2016, 2019; Otálora et al. 2013a).

The generated sequences were assembled and edited using Geneious version R8 (http://www.geneious.com, Kearse et al. 2012). All edited sequences underwent initial identity verification through BLAST searches (Zhang et al. 2000). The alignment of these sequences was performed using the MUSCLE algorithm (Edgar 2004) in AliView 1.09 (Larsson 2014). An intron in the b-tub was excluded. Indels in the mtSSU were not deleted, allowing smaller gap positions within the final block. The five genetic regions defined above were analysed separately using Maximum Likelihood (ML). As no significant incongruence was detected, the alignments were concatenated. The final alignment has been deposited in TREEBASE (http://www.treebase.org) with accession number (http://purl.org/phylo/treebase/phylows/study/TB:S31975). After concatenation, we inferred phylogenetic relationships using ML with the same settings used as in the individual gene analyses utilizing RAxML. Likelihood and ML bootstrapping were executed through RAxML 8 (Stamatakis 2014) implementing a general time reversible (GTR) model of nucleotide substitution with gamma distributed rate heterogeneity (GTRGAMMA). One thousand bootstrap (BS) replicates were completed using the non-parametric BS algorithm of RAxML–HPC v.8 on XSEDE using the CIPRES Web Portal (Miller et al. 2010).

For the final concatenated dataset, we included 131 nucleotide sequences of mtSSU, b-tub, MCM7, and the two parts of RPB2 (5–7 and 7–11). This dataset encompassed 4140 nucleotide positions (843 bp for mtSSU, 708 bp for b-tub, 597 bp for MCM7, and 1074 bp for RPB2 5–7 and 918 bp for RPB2 7–11) with 39 terminals. It incorporates representatives from selected genera within the Collemataceae, in conjunction with additional data sourced from GenBank (utilized in Otálora et al 2013a, 2013b; Košuthová et al. 2019), with Enchylium bachmanianum as the outgroup. None of the sequences of the “Nigrescens” and “Flaccidum”-groups previously used in phylogenetic analyses and uploaded to GenBank (Otálora et al. 2010, 2013a, 2013b; Miadlikowska et al. 2014), were included in this study due to concerns about specimen misidentification. Sequences used in this study are summarized in Fig. 1, Table 2.

Figure 1. 

The most likely tree from RAxML analysis based on 4140 aligned characters of mtSSU, b-tub, MCM7 and RPB2 5–7 and 7–11 from 39 specimens. Support values (Likelihood BS) given when BS ≥ 60%. The scale bars indicate 0.03 substitution.

We generated 104 new sequences (Table 2), comprising 28 mtSSU, 23 b-tub, 16 MCM7, 27 RPB2. Our analyses resulted in a topology (Fig. 1) very similar to the findings of Otálora et al. (2013a, 2013b).

Our phylogenetic analysis reveals that European Collema should be treated as including C. curtisporum, C. flaccidum, C. furfuraceum, C. nigrescens, C. ryssoleum, C. subflaccidum and C. subnigrescens, but excluding C. glebulentum as this species clearly groups within Leptogium together with Leptogium azureum, the conserved type species of Leptogium (Fig. 1). The groups within Collema as informally circumscribed by Degelius (1954), are not supported by our phylogeny. The “Flaccidum”-group, consisting of C. flaccidum and C. subflaccidum, is nested within the “Nigrescens” -group. This indicates that these two informal groups are not useful for a phylogenetically based classification (Fig. 1). Collema furfuraceum is further non-monophyletic, suggesting the need for further taxonomic revision.

Degelius (1954) differentiated his informal groups in European Collema based on morphological characteristics, noting that species within the “Flaccidum”-group are all isidiate but not pustulate (Table 1, Fig. 2A–F). Isidia in Collema flaccidum are typically numerous and squamiform (flattened) when fully developed, rounded, and about 0.2–0.5 mm broad or larger. Larger isidia are often crenate or lobulate, with occasional teretiform isidia mixed in (Fig. 2B). In contrast, C. subflaccidum has smaller, globular isidia approximately 0.05–0.1 mm in diameter that become slightly teretiform as they mature, reaching at least 0.3 mm in length, with simple or slightly branched forms (Fig. 2C, D). Collema glebulentum (Fig. 2E, F) has, sometimes together with a primitive pseudocortex, a distinct typical pseudocortex, often developed on the lower surface, especially on smaller lobes. When a typical pseudocortex is present, it is composed of several cell layers where the cells can be quite large, reaching up to 15 µm in diameter (Fig. 3A). Parts of the thallus may have an entirely paraplectenchymatous structure. This characteristic, noted by Degelius (1954), actually supports the original classification of C. glebulentum in Leptogium (Fig. 3B). In areas where the thallus is not entirely paraplectenchymatous, its structure resembles that of other species in the “Flaccidum” and “Nigrescens”-groups. The entire thallus in species from these groups is composed of hyphae that are either loosely or compactly interwoven, or arranged distinctly perpendicular to the upper and lower cortices throughout the thallus (Fig. 3C).

Figure 2. 

Thallus habitus and isidia of A Collema flaccidum (UPS-L872188) B C. flaccidum (UPS-L872188) C C. subflaccidum (S-F492349) D C. subflaccidum (S-F492350) E Leptogium glebulentum (S-L49768) F L. glebulentum – dwarf form (S-L49768) G C. ryssoleum (UPS-L933969) H C. ryssoleum (UPS-L933969) I C. subnigrescens (S-F57732) J C. curtisporum (UPS-L111603) K C. nigrescens (UPS-L934034) L C. nigrescens (UPS-L934034) M C. furfuraceum (UPS-L934040) N C. furfuraceum (UPS-L934040). is = isidia, ap = apothecium. Scale bar: A–C, E–N: 1 mm, D: 0.5 mm.

Figure 3. 

Thalli transversal cross-sections in water A–C, ascospores D–G A thallus with typical pseudocortex (Leptogium glebulentum S-L49768) B thallus paraplectenchymateous throughout (Leptogium glebulentum S-L49768), C thallus with hyphae which are perpendicular to the surface (C. furfuraceum S-F492354) D spore of C. ryssoleum (UPS-L178905) E spore of C. subnigrescens (GZU66-201) F spore of C. curtisporum (UPS-L111603) G C. nigrescens (S-F492352), tPsC = typical pseudocortex, Hp = hyphae, pPlect = paraplechtenchyma. Scale bar: 10 μm.

Species within the “Nigrescens”-group are characterized by their pustulate and ridged thalli (Fig. 2G–P) and can be divided into non-isidiate species and those that produce isidia. Among the non-isidiate species, Collema ryssoleum (Fig. 2G, H) is distinct due to its spore morphology. The spores are short, measuring up to 40 µm in length, similar to those in C. curtisporum. However, unlike others in the group, C. ryssoleum spores are relatively wide (up to 8.5 µm) with acute ends (Table 1). Additionally, this species exhibits a unique Mediterranean distribution in the temperate zone and is adapted to a saxicolous habitat (Fig. 3D). Collema subnigrescens is characterized by an up to 20 cm large thallus (Fig. 2I) and by its narrowly fusiform to irregularly clavate spores with twisted, acute ends (5-celled, occasionally up to 12 cells; Fig. 3E). It closely resembles C. curtisporum, which, however, has a smaller, up to 4 cm wide thallus (Fig. 2J) and shorter, mostly 3-celled bacillariform spores with obtuse ends and a typical “curved-acute shape” (Fig. 3F). Spores in C. nigrescens differ from the other species in being acicular to bacillariform, longer and thinner than those in C. subnigrescens (usually around 60 µm long and less than 5 µm wide), 6–13 celled (Fig. 3G).

Among the isidiate species, Collema nigrescens is notable for its globular isidia (ca 0.2 mm in diameter; Fig. 2K, L), whereas C. furfuraceum has thinner, teretiform isidia (ca 0.05–0.15 mm wide) that become coralloid and reach up to 0.3 mm in length (Fig. 2M, N). Degelius (1974) later included Collema luzonense Räs. from the Philippines as a variety of C. furfuraceum and the var. luzorense, was distinguished by its euparaplectenchymatous excipulum proprium and frequent presence of white-pruinose apothecia. Another distinguishing feature of var. luzorense is its thicker lobes, reaching up to 300 µm, compared to 100 µm in var. furfuraceum. The var. luzorense has not been reported from Europe and the thallus thickness of the samples (which were all sterile) in our study suggests that they belong in var. furfuraceum.

Given the results of our phylogenetic analysis, it is crucial to study C. furfuraceum further. This study is under way, whereby we also plan to investigate the potential role of isidia as a distinguishing feature for species identification.

The authors have declared that no competing interests exist.

No ethical statement was reported.

We received generous funding from the Swedish Taxonomy Initiative (Svenska Artprojektet) administered by the Swedish Species Information Center (ArtDatabanken; grants 2016-207 4.3 and SLU.dha.2019.4.3-48), and the Stenholms donationsfond (GBT 2018-00002-26).

Conceptualization: MW, AK. Data curation: AK. Formal analysis: AK. Funding acquisition: AK, MW. Investigation: UN, FJ. Methodology: AK. Validation: FJ, UN, MW. Visualization: UN. Writing - original draft: AK. Writing - review and editing: UN, FJ, MW.

Alica Košuthová https://orcid.org/0000-0001-5991-7444

Mats Wedin https://orcid.org/0000-0002-8295-5198

All of the data that support the findings of this study are available in the main text.