Phylogenetic study of the Cladonia cervicornis group (Cladoniaceae, Lecanorales) discloses a new species, Cladonia teuvoana

Abstract

The Cladonia cervicornis group comprises lichen-forming fungi characterized by having scyphi with central proliferations. It includes c. 20 species globally. The taxonomy of this group is poorly resolved, with many species not thoroughly disentangled. The focus of this study is the European species in the C. cervicornis group. In order to estimate the phylogenetic relationships of these species, six loci were used: ITS rDNA, IGS rDNA, RPB1, RPB2, ef1α and cox1. Species delimitation methods (ASAP, PTP and GMYC) were used to infer the species boundaries based on four loci, ITS rDNA, IGS rDNA, cox1 and RPB2. A morphological analysis based on multivariate methods was performed to assess the importance of phenotypic differences among the lineages. The phylogenetic reconstructions placed the species of this group in the subclade Cladonia. Five lineages were recovered, corresponding to C. cervicornis, C. macrophyllodes, C. pulvinata, C. verticillata and a new lineage that we describe here, C. teuvoana. Our analyses revealed that Cladonia cineracea, C. stricta and C. trassii are polyphyletic.

Introduction

The genus Cladonia is one of the most speciose genera of macrolichens (Lücking et al. 2017), with more than 475 accepted species described (Stenroos et al. 2019). Most of the species are terricolous, growing in open areas that have a certain amount of moisture (Ahti 2000). The genus has a subcosmopolitan distribution. It is notably diverse in the Neotropics (c. 195 species), particularly Eastern Brazil, and Australasia-Melanesia with around 40% of species endemic for this region (Ahti & Aptroot 1992; Ahti 2000; Aptroot et al. 2021), while in Europe c. 105 species have been reported (Litterski & Ahti 2004). The genus Cladonia is easily recognizable by usually having a dimorphic thallus, consisting of a squamulose or crustaceous primary thallus and a fruticose secondary thallus. However, species identification within this genus can be challenging (Burgaz et al. 2020; Černajová et al. 2022). Most of the taxonomic characters used to distinguish Cladonia species are associated with the secondary thallus, called podetium (Ahti 2000). Due to the limited number of characters available, distinguishing species based on squamule morphology is often difficult (Burgaz et al. 2020). However, many species have a dominant primary thallus and rarely develop podetia (Ahti 2000; Pino-Bodas et al. 2010a, 2012). In these cases, morphological characters associated with the primary thallus are indispensable for distinguishing species. The study of secondary metabolites can also be of great use for species circumscription (Huovinen & Ahti 1982; Culberson 1986), since morphologically similar species sometimes contain different secondary metabolites, for example C. subulata (L.) F. H. Wigg. and C. rei Schaer. or C. humilis (With.) J. R. Laundon and C. conista (Nyl.) Robbins (Dolnik et al. 2010; Pino-Bodas et al. 2010b, 2013a).

This study focuses on the C. cervicornis (Ach.) Flot. group in Europe. Members of this group are characterized by scyphose podetia with a corticate surface and by having proliferations arising from the centre of the scyphi. There are c. 20 species of Cladonia globally with these features (Ahti 2007). The most recent phylogeny of the genus placed them within the clade Cladonia, subclade Cladonia. This clade also includes species of the former sections Cladonia, Helopodium and Unciales, including the type species of the genus, C. subulata (Stenroos et al. 2019). Numerous authors have paid attention to the taxonomy of species with central proliferations in different geographical regions (Abbayes 1949; Ahti & Marcelli 1995; Ahti 1978, 1980, 1983, 1998, 2007; van Herk & Aptroot 2003; Pino-Bodas et al. 2010a; Charnei & Eliasaro 2013). However, many taxa are still not well understood (Ahti 2000, 2007) and phylogenetic studies have shown that some are polyphyletic (Pino-Bodas et al. 2013b; Stenroos et al. 2019).

Nine species in the C. cervicornis group have been previously reported from Europe: C. cervicornis, C. firma (Nyl.) Nyl., C. macrophyllodes Nyl., C. pulvinata (Sandst.) Van Herk & Aptroot, C. stricta (Nyl.) Nyl., C. subcervicornis (Vain.) Kernst., C. trassii Ahti, C. uliginosa (Ahti) Ahti, and C. verticillata (Hoffm.) Schaer. (Ahti & Stenroos 2013). However, phylogenetic studies have shown that C. firma and C. subcervicornis do not belong to this group as they are phylogenetically distant from the rest of the species (Stenroos et al. 2019). In addition, there is another species restricted to Macaronesia, C. microphylla Ahti & Aptroot (Ahti & Aptroot 2009). Cladonia cervicornis, C. pulvinata and C. verticillata have a wide distribution in Europe (van Herk & Aptroot 2003; Ahti & Stenroos 2013; Burgaz et al. 2020; Pino-Bodas et al. 2021), while C. stricta, C. trassii and C. uliginosa are primarily distributed in the arctic or arctic-boreal regions of Eurasia and North America, with some outposts in the Alps, although their distribution is not yet well known (Ahti 1998; Brodo et al. 2001; Ahti & Stenroos 2013; Pino-Bodas et al. 2021). Cladonia macrophyllodes is more widely distributed, but outside arctic regions it is restricted to mountainous areas (Goward & Ahti 1997; Ahti & Stenroos 2013; Burgaz et al. 2018, 2020). The distribution patterns of C. cervicornis, C. pulvinata and C. verticillata differ, although in several regions of Europe they often coexist, and morphological discrimination can be difficult (Ahti 1980; Ahti & Stenroos 2013; Søchting 2017; Gheza et al. 2018; Burgaz et al. 2020). Morphological studies by van Herk & Aptroot (2003) showed that the coloration and morphology of the primary thallus can be used to distinguish these three taxa. In addition, the presence of psoromic acid allows C. pulvinata to be distinguished from the other two species, which do not synthesize this compound. Cladonia cervicornis is a highly variable species and some authors have indicated the need for further studies to clarify its taxonomy (Ahti & Stenroos 2013; Burgaz et al. 2020). Specimens of these taxa have been the subject of some molecular research (Pino-Bodas et al. 2010a, 2013b; Osyzcka et al. 2018; Stenroos et al. 2019), the results of which indicate that C. cervicornis, C. pulvinata and C. verticillata form independent lineages. However, it should be noted that sampling has been mostly focused on the Mediterranean region, where C. cervicornis is particularly common (Burgaz et al. 2020), while less sampling has been carried out in Central and Northern Europe where C. verticillata is more frequent and more difficult to distinguish from C. cervicornis (Ahti & Stenroos 2013). Phylogenetic studies have shown C. verticillata to be polyphyletic (Pino-Bodas et al. 2013b; Stenroos et al. 2019), and more extensive sampling will be necessary to determine whether European specimens belong to a single species.

The aims of this study were: 1) to evaluate the monophyly of the species within the C. cervicornis group reported for Europe; 2) to estimate the phylogenetic relationships of the species of the C. cervicornis group within the subclade Cladonia; 3) to carry out a morphological study that characterizes each of the lineages found.

Material and Methods

Taxon sampling

This study is based on 160 specimens from 18 countries. Specimens representing all known European species with central proliferations were included in the phylogenetic analyses. New study specimens included: C. cervicornis (n = 57), C. cineracea Ahti (n = 1), C. macrophyllodes (n = 26), C. microphylla (n = 7), C. pulvinata (n = 15), C. stricta (n = 4), C. trassii (n = 6), C. uliginosa (n = 2) and C. verticillata (n = 32). In addition, a selection of specimens from other regions of the world were included. A small number of specimens of C. rappii A. Evans, a species difficult to distinguish from C. verticillata in North America, were newly sequenced (Table 1). The specimens studied are deposited in C, H, E, K, MACB and TROM. Types of C. cervicornis (lectotype H-ACH 1672B and isolectotype BM-ACH 722A), C. macrophyllodes (H-ACH 1673), C. pulvinata (lectotype H), and representative specimens of C. sobolifera Nyl. (H-NYL-0608, 0638, 38780, 38784, 38785, 387986) were studied to compare their morphological features with our specimens.

Table 1. List of subclade Cladonia specimens included in the study with voucher information, secondary metabolites detected by thin-layer chromatography and GenBank accession numbers. New sequences are in bold.

ATR = atranorin; FUM = fumarprotocetraric acid; CNSTIC = connorstictic acid; CPH-2 = confumarprotocetraric acid; CPSO = conpsoromic acid; NSTIC = norstictic acid; PHY = physodalic acid; PSO = psoromic acid; USN = usnic acid; ZEO = Zeorin

Phenotypic study

The secondary metabolites of each specimen were studied by thin-layer chromatography (TLC). The standard protocol was used as described in White & James (1985) and Orange et al. (2010), with solvents A (toluene:1,4-dioxane:acetic acid) and C (toluene:acetic acid). The specimens were identified following van Herk & Aptroot (2003), Ahti (2007), Ahti & Stenroos (2013) and Burgaz et al. (2020). The morphological study was focused on C. cervicornis, C. pulvinata and C. verticillata which are the more common taxa in Europe and difficult to distinguish from each other. Based on previous studies (van Herk & Aptroot 2003; Ahti 2007; Ahti & Stenroos 2013; Burgaz et al. 2020), 13 quantitative traits were selected to conduct the morphological analyses: length of squamules, width at the middle of squamules, maximum width of squamules, squamule incision depth, length of the blackened base of squamules, thickness of squamules, length of first scyphi, total length of podetium, number of proliferations, number of tiers, widening in lower half of podetium, width of scyphi, and thickness of podetial wall. To better represent morphological patterns, the length ratio of the variables ‘width of squamules’, ‘width of scyphi’, ‘squamules incisions depth’ and ‘blackened base of squamules’ were used. To determine whether the widening of the scyphi is gradual from the base or abrupt above the middle of the podetium, the difference in width of the podetium was analyzed, corrected for length (width at the middle – width at the base / 1/2 (podetium length)). These traits were examined in a selection of 63 specimens, representing all phylogenetic lineages, by measuring three squamules and three podetia per specimen (whenever the amount of material permitted) under a dissecting microscope (Nikon SMZ21000) using a digital caliper. In addition to quantitative characters, the following qualitative features were also studied: colour of the underside of the squamules, colour of the upper side of the squamules, morphology of the squamules, morphology of the squamule lobes, presence of veins on the underside of the squamules, presence of rhizines and continuity of the podetial algal layer.

All statistical analyses were conducted in R v. 4.1.2 (R Core Team 2022). The normality and homoscedasticity of the quantitative variables were assessed using the Shapiro-Wilk and Bartlett tests, respectively. It was found that only ‘depth of incisions’ was normally distributed. Due to the lack of normality, the variables did not meet the assumptions of a principal component analysis (PCA), so the dataset was analyzed using non-metric multidimensional scaling (NMDS; Kruskal 1964), a more robust method for this type of data (Austin 1985; Minchin 1987). NMDS was conducted with the Bray-Curtis distances in the vegan package (Oksanen et al. 2013). Permutational multivariate analysis of variance (PERMANOVA) was used to test differences between clades and species, with the Bray-Curtis dissimilarity, 999 permutations and Bonferroni correction for multiple comparisons.

Contingency tables were used to assess the association between qualitative characters and different clades. The morphological characters studied were: upcurled dry squamules, upper surface colour, lanceolate lobes, colour of base squamules, continuous algal layer and basal colour of podetia (alternative states are listed in Supplementary Material Table S1, available online). The analyses were performed using the abind package (Plate & Heiberger 2016). As the number of specimens in clade C was low, Fisher's test was used instead of the chi-square test (Agresti 2018).

DNA extraction, PCRs and sequencing

A single podetium per specimen was used for DNA extraction. Exceptionally, for the molecular biology study, specimens with poorly developed podetia or without them were included. Before DNA extraction, the fragment was soaked in acetone for 1 h to remove secondary metabolites. The acetone extracts were used to conduct TLC. Total genomic DNA was extracted using the E.Z.N.A. Forensic Kit (Omega) following the manufacturer's instructions; DNA was dissolved in 100 μl of elution buffer (10 mM Tris-Cl). Based on preliminary studies (Pino-Bodas et al. 2010a, 2013a, c; Stenroos et al. 2019), the regions ITS rDNA, IGS rDNA, RPB2 and cox1 were selected to conduct the revision of the C. cervicornis group. The primers and PCR conditions are described in Pino-Bodas et al. (2010b) and Stenroos et al. (2019). PCR reactions were carried out using BIOTAQ polymerase (Bioline), including 0.3 μl of Taq polymerase, 2.5 μl of 10× PCR buffer, 1.4 μl of MgCl₂ (50 μm μl⁻¹), 1.6 μl of dNTPs (2.5 μm μl⁻¹), 1 μl of BSA (1 μm μl⁻¹), 1 μl of each primer (10 μm μl⁻¹), and 1 μl of extracted DNA. PCR products were sequenced at Macrogen Spain (http://www.macrogen.com).

Phylogenetic analyses

The dataset used for the phylogenetic analyses included the specimens newly sequenced from the C. cervicornis group, sequences of Cladonia species with central proliferation from other regions of the world included in clade Cladonia, subclade Cladonia, and other species without central proliferations but belonging to this clade (Stenroos et al. 2019). Additionally, specimens of Cladonia species representing other major lineages within the clade Cladonia (Stenroos et al. 2019) were included in the phylogenetic analyses (Table 1). Cladonia boryi Tuck. and C. nipponica Vain., belonging to clade Boryia, were used as an outgroup. The sequences were assembled using Sequencher v. 4.1.4 (Gene Codes Corporation, Inc., Ann Arbor, Michigan, USA). The alignments of each genetic region were performed using the online version of MAFFT (Katoh & Standley 2013), with default parameters. Ambiguous positions of ITS rDNA and IGS rDNA alignments were removed with Gblocks v. 0.91b (Castresana 2000) using less restricted options (allowing final blocks, gap positions and less strict flanking positions).

Maximum likelihood (ML) analyses were performed in RAxML v. 7.04 (Stamatakis et al. 2008) for each locus, using 500 pseudoreplicates of rapid bootstrap and the GTRGAMMA model. The congruence among loci was inspected manually, based on the method described by Kauff & Lutzoni (2002, 2003). No incongruences were detected, and the datasets were combined. In addition to the genetic regions amplified in the specimens of the C. cervicornis group, concatenated analyses also included RPB1 and ef1α loci, which were used to estimate the phylogeny of Cladonia (Stenroos et al. 2019) and numerous sequences for specimens of the Cladonia subclade were available. This approach increased the number of loci sampled within the clades in the phylogeny. We verified that RPB1 and ef1α were not represented exclusively in specimens of some branches of the phylogeny. The concatenated dataset was restricted to specimens represented by at least two loci. The concatenated dataset was analyzed by ML and Bayesian inference. JModelTest (Posada 2008) under the Akaike information criterion (AIC) was used to select the best substitution model for each region. The selected models were: TrNef + G for ITS rDNA, IGS rDNA and RPB2; K80 + G for cox1; SYM + G for ef1α and TrNef + I + G for RPB1. The partitioning scheme used to analyze the concatenated datasets considered 15 partitions: ITS rDNA, IGS rDNA and each codon position for the loci cox1, ef1α, RPB1 and RPB2. ML analysis was conducted in RAxML using the GTRGAMMA model for each partition and 1000 pseudoreplicates of rapid bootstrap. The Bayesian analyses were performed using MrBayes v. 3.2.6 (Ronquist et al. 2012). Posterior probabilities were approximated by sampling trees using Markov chain Monte Carlo (MCMC). Two simultaneous runs, each with 20 000 000 generations and employing four chains, were executed. Every 1000th tree was saved into a file. The first 25% of trees was discarded as burn-in. Convergence between chains was assessed using Tracer v. 1.7 (Rambaut et al. 2018). The likelihood versus generation number and the average standard deviation of split frequencies (≤ 0.01) were plotted. Branches with posterior probabilities ≥ 0.95 and bootstrap values ≥ 70% were considered strongly supported. The phylogenetic analyses based on the concatenated dataset were performed on the CIPRES Science Gateway web portal (Miller et al. 2010).

Species delimitation analyses

In addition to the clade delimitation obtained in the phylogenetic reconstructions, three species delimitation methods were employed to assess the number of species within the C. cervicornis group: Automatic Partitioning (ASAP) (Puillandre et al. 2021), Poisson Tree Processes (PTP) (Zhang et al. 2013) and General Mixed Yule Coalescent (GMYC) (Pons et al. 2006). These algorithms belong to the category of ‘discovery’ species delimitation methods, since they do not require any prior information on species boundaries. These analyses included only putative species represented by more than two sequences, removing singletons that may create artefacts (Lim et al. 2012; Pentinsaari et al. 2017). ASAP was implemented on the web server (https://bioinfo.mnhn.fr/abi/public/asap/) using the Jukes-Cantor (JC69) model to compute the genetic distances. PTP analyses were performed using the best ML tree estimated in RAxML for each locus. The analyses were conducted on the webserver (http://species.h-its.org/) with default parameters that included 500 000 generations, with a thinning rate of 100 and 10% burn-in. GMYC was conducted using iTaxoTools (Vences et al. 2021), using a single threshold. ML trees of each locus were transformed into ultrametric trees with the ‘chromos’ function of the ape package (Paradis & Schliep 2019) for R v. 4.3.3 (http://www.rproject.org/) and used for GMYC analyses.

Results

In this study, 469 new DNA sequences were generated (119 of ITS rDNA, 145 of IGS rDNA, 95 of RPB2 and 110 of cox1). The concatenated dataset was composed of 221 specimens and 3798 characters. The ML analysis resulted in a tree with −Lnl = 27581.920. The Bayesian analyses produced a consensus tree with an arithmetic mean of −Lnl = 102147.35. ML and Bayesian consensus trees showed similar topologies. Figure 1 shows the Bayesian 50% majority-rule consensus tree, which is consistent with the topology of the subclade Cladonia generated in the phylogeny of Stenroos et al. (2019). The subclade Cladonia was revealed as monophyletic with high support, while the centrally proliferating species are not monophyletic. All specimens of C. verticillata from Europe form a monophyletic clade, although this is supported only in the Bayesian analysis (PP = 1/Bootstrap = 64). A small number of specimens from North America are included in this clade, which is closely related to C. subulata. The results appear to suggest that C. trassii is polyphyletic; one specimen of C. trassii is basal to the clade formed by C. verticillata and C. subulata, while three other specimens occur distantly, close to a specimen of C. cineracea and C. rappii. Cladonia pulvinata was monophyletic and highly supported in both analyses. Specimens identified as C. cervicornis form two well-supported clades. One clade (clade C), with seven specimens, is closely related to a clade comprising C. ceratophyllina (Nyl.) Vain., C. centrophora Müll. Arg., C. andesita Vain. and C. nitens Ahti. Clade D includes most of the specimens of C. cervicornis in addition to all the specimens studied of C. microphylla. Cladonia macrophyllodes is found to be monophyletic and closely related to C. uliginosa, which is represented by two specimens in the phylogeny. One specimen (Lendemer 22968) morphologically related to C. cervicornis from North America appears to be related to C. mateocyatha Robbins.

Figure 1. Phylogeny of subclade Cladonia, based on concatenated ITS rDNA, IGS rDNA, RPB1, RPB2, ef1α and cox1, showing the 50% majority-rule consensus tree of the Bayesian analysis. Branches with a bootstrap value of > 70% and a posterior probability of > 0.95 are in bold. Species of the C. cervicornis group are indicated in different colours. The country of origin of the specimen is indicated in brackets using the ISO abbreviations. In colour online.

The results of species delimitation methods are presented in Fig. 2. The number of inferred species ranges from 2 to 115. In general, these methods showed little congruence with each other and with the current species delimitation. However, several of the methods for different loci identified clade C of C. cervicornis as a different species. Specifically, this putative species was inferred by ASAP for ITS rDNA, RPB2 and cox1, by PTP for ITS rDNA and RPB2, and by GMYC for RPB2.

Figure 2. Results of species delimitation analyses (Automatic Partitioning (ASAP) (Puillandre et al. 2021); Poisson Tree Processes (PTP) (Zhang et al. 2013); General Mixed Yule Coalescent (GMYC) (Pons et al. 2006)) in the Clandonia cervicornis group, based on ITS rDNA, IGS rDNA, cox1 and RPB2. The same colour indicates that the specimens were inferred to belong to the same species. The absence of colour indicates that the specimens were not analyzed. The DNA code corresponds to that shown in Fig. 1. Further details of the specimens can be found in Table 1. In colour online.

Table 1 shows the secondary metabolites detected by TLC in each specimen studied. The results show a high chemical homogeneity within the taxa. As a major substance, all the specimens of C. pulvinata contained psoromic acid, and all the specimens of C. verticillata contained fumarprotocetraric acid. Most C. cervicornis samples contained fumarprotocetraric acid, except for a small number of samples from Cyprus which also contained atranorin. Traces of other substances such as physodalic acid, confumarprotocetraric acid and a fatty acid with an R _f compatible with rangiformic acid were detected in specimens from different geographical origins (Table 1).

Table 2 presents the PERMANOVA pairwise comparison. PERMANOVA analyses showed a clear morphological differentiation between C. cervicornis and C. verticillata, and between C. cervicornis and C. pulvinata. However, there were no significant differences between C. verticillata and C. pulvinata. Our analyses showed that the specimens of clade C were morphologically very similar to C. pulvinata. These results were confirmed by NMDS analysis (Fig. 3), where specimens of C. cervicornis and C. verticillata clearly form separate clusters, while those of C. pulvinata and clade C overlap.

Table 2. PERMANOVA results of morphological pairwise comparison among subclade Cladonia cervicornis group. Significant P-values are in bold.

Figure 3. Non-metric multidimensional scaling (NMDS) plot, illustrating the morphological variation among the species belonging to the Cladonia cervicornis group. Stress = 0.14. In colour online.

The contingency table analyses showed significant associations between the clades and the characters: lower surface colour (P-value = 0.0001034), upper surface colour (P-value = 0.01213), lanceolate/rounded lobes (P-value = 0.001965), basal colour of squamules (P-value = 1.917e-11), and basal colour of podetia (P-value = 0.01323), although there were no significant associations among clades for continuous/discontinuous algal layer (P-value = 0.11).

Discussion

Species delimitation of the Cladonia cervicornis group in Europe

It is well known that many species of Cladonia present a high level of morphological intraspecific variation. This variation may be due to different stages of development, to changes produced by different environmental conditions, such as different light intensity or humidity, or a response to anthropogenic disturbances (Osyzcka & Rola 2013; Osyczka et al. 2014, 2018; Pino-Bodas et al. 2015). Therefore, determining which phenotypic traits can be used as diagnostic characters is often challenging. Species delimitation in Cladonia has been hampered by the low genetic variability of the genus, which means that the markers used in phylogenetic and species delimitation studies in fungi often do not have sufficient resolution (Pino-Bodas et al. 2011, 2013c; Stenroos et al. 2019). Several studies have shown that phenotypically and ecologically disparate species cannot be genetically distinguished even using multiple loci (Stenroos et al. 2019; Steinová et al. 2022). The low genetic variation among Cladonia species may be attributed to recent divergence among species or to the dominance of asexual reproduction (Ahti 2000; Stenroos et al. 2019; Steinová et al. 2022).

This research investigates the morphological and genetic variation of the C. cervicornis group in Europe, focusing on the three most common species, C. cervicornis, C. pulvinata and C. verticillata. Characters traditionally used to distinguish species of the C. cervicornis group have been: the number of central proliferations, the number of tiers, the presence of a melanotic region at the base of the podetia, gradual or abrupt podetial tapering, morphology of squamules, and the secondary metabolites (Ahti 2000, 2007; van Herk & Aptroot 2003). Although preliminary phylogenetic studies indicated that these three taxa were monophyletic and constituted independent lineages (Pino-Bodas et al. 2010a, 2013b; Stenroos et al. 2019), the large morphological variation documented in C. cervicornis and the difficulty in distinguishing some specimens from C. verticillata (Ahti & Stenroos 2013; Søchting 2017; Burgaz et al. 2020) prompted us to perform a new phylogenetic study. In this study, we used an integrative approach, combining the results of phylogenetic analyses, different methods for molecular species delimitation and the phenotypic study of specimens, to address species delimitation in the C. cervicornis group in Europe. Phylogenetic analyses showed five clades (Fig. 1), one corresponding to C. verticillata (clade A), another to C. pulvinata (clade B), two clades including specimens previously identified as C. cervicornis (clades C and D) and one corresponding to C. macrophyllodes (clade E). Nevertheless, these five clades were not identified as distinct species by the different species delimitation methods employed (Fig. 2). In fact, the results of the different species delimitation methods displayed great inconsistencies between them (Fig. 2), most showing an unrealistic amount of oversplitting. Although species delimitation methods have been widely used to establish species boundaries more objectively within different groups of lichen-forming fungi (Parnmen et al. 2012; Del-Prado et al. 2016; Pérez-Ortega et al. 2016; Wei et al. 2016; Košuthová et al. 2020), it has been common to obtain inconsistent or unrealistic results in terms of the number of species delimited (Pino-Bodas et al. 2018; Mercado-Díaz et al. 2020; Myllys et al. 2023). Inconsistencies may be attributed to breaking assumptions of different methods, employing low numbers of specimens, lack of sampling across the geographical range of species, or recent divergence (Carstens et al. 2013; Magoga et al. 2021). Although our sampling was extensive, with a large number of specimens analyzed, it was focused on Europe and might be geographically biased for some lineages. Reciprocal monophyly has been used as a criterion for species delimitation, since it is indicative of sufficient divergence between lineages (De Queiroz 1998; Lücking et al. 2021). Based on this criterion and our knowledge of genetic variation in Cladonia, we consider that the lineages found in the phylogenetic analyses represent different species. In addition, most of these lineages are supported by phenotypic differences. Our morphological analyses not only document variation in each lineage, but also show morphological differences among most of them (Fig. 3, Table 2), confirming the results of van Herk & Aptroot (2003).

Our analyses show that all European C. verticillata specimens belong to a single lineage, clade A. Therefore, this clade represents C. verticillata s. str. Furthermore, morphological analyses indicate morphological differences with the specimens of C. cervicornis. Clade A not only included European specimens, but also some from North America. Within this clade, two well-supported subclades could be distinguished, one containing three specimens from North America, the other with all European specimens and some from North America. However, these subclades were not delimited as independent species by any of the species delimitation methods. Other specimens of C. verticillata from other regions were not included in this clade, which confirms that a worldwide study is required to clarify the taxonomy of C. verticillata (Ahti 2007; Pino-Bodas et al. 2021).

Clade B consists exclusively of C. pulvinata specimens. According to our morphological analyses, the distinction between C. pulvinata and C. verticillata is difficult, since they display a large amount of overlap (Fig. 3). However, in regions where the species grow side-by-side, they can usually be reliably distinguished (van Herk & Aptroot 2003), and the presence of psoromic acid in C. pulvinata can always be used as a diagnostic character to identify the specimens of this lineage.

Clade C encompasses specimens with a more or less Atlantic distribution. Morphological analyses indicate differences from the rest of the lineages of the C. cervicornis group. This lineage was also supported by some species delimitation analyses (Fig. 2). From a morphological point of view, this lineage is distinguished by the narrower scyphi compared to other taxa within the group. These scyphi display abrupt widening and no tier proliferations have been observed. The squamules are smaller than those of the other lineages and may have a blackened zone on the underside, reaching up to 1/3 of the length of the squamules, which are white at the ends. Study of the type material has allowed us to determine that C. cervicornis s. str. corresponds to clade D. The primary thallus of the C. cervicornis type (lectotype and isolectotype) is morphologically similar to that of the specimens of this clade (Fig. 4), characterized by large, deeply incised squamules and an underside with purplish-greyish tinged ends. Possible designations applicable to clade C would be C. cristata Hoffm. or C. sobolifera. However, neither the iconography presented by Dillenius (1742) nor the description by Hoffmann (1796) matches the morphology of the specimens in this clade and we therefore rule out applying C. cristata to this clade. We did not have access to the type of C. sobolifera but we studied the Nylander specimens in the Helsinki herbarium and conclude that they belong to C. cervicornis (Fig. 4). Therefore, we consider clade C to be a new species that we describe here under the name Cladonia teuvoana.

Figure 4. A–C, Cladonia teuvoana sp. nov. (MACB 127525—holotype). D, Cladonia sobolifera (H-NYL 0608). E, Cladonia cervicornis (BM-ACH 722A—isolectotype). Scales: A, B & D = 1 cm; C = 0.5 cm. In colour online.

Clade D consists of the specimens of C. cervicornis s. str. together with all the analyzed specimens of C. microphylla (Fig. 1), a Macaronesian endemic species characterized by deeply divided squamules (Ahti & Aptroot 2009). Although the analyses performed here did not distinguish this species from C. cervicornis, we do not propose any taxonomic changes since we did not undertake an in-depth morphological study of this taxon. In addition, phylogenomic analyses have resolved relationships between closely related species of lichen-forming fungi that conventional markers (e.g. ITS rDNA, LSU rDNA, RPB2) used in fungal phylogenies failed to separate (Grewe et al. 2018). We expect that analyses at the genomic scale will provide a better understanding of the boundaries between C. cervicornis and C. microphylla. Cladonia cervicornis specimens grouped in this clade show a high morphological variation (Fig. 3), but this variation is not related to genetic variation. Despite the low genetic variation found, several of the species delimitation methods divided the clade into several species. Given the lack of congruence among the delimitation methods, we opted for the most conservative option, which is to keep this lineage as a single species, a hypothesis that was supported only by ASAP for ITS rDNA and RPB2. Cladonia cervicornis has been reported from North America (Hammer & Ahti 1990; Hammer 1995; Ahti & Hammer 2002) but the identity of these specimens is uncertain. Our results, based on the study of a single specimen, indicate that C. cervicornis s. str. is probably not present in North America (Fig. 1).

Clade E includes all the analyzed specimens of C. macrophyllodes. No circumscription problems have been reported for this species; although the primary thallus could be confused with that of C. subcervicornis, the two species have somewhat different habitat requirements (Ahti & Stenroos 2013; Burgaz et al. 2020). However, as for the remaining lineages, most species delimitation analyses inferred more than one species within this lineage, with the exception of ASAP and PTP for cox1. In the absence of other evidence, we maintain C. macrophyllodes as a single species.

Phylogenetic relationships within the Cladonia cervicornis group

The phylogenetic placement of species with central proliferations has been examined in previous studies (Stenroos et al. 2002, 2019; Beiggi & Piercey-Normore 2007; Pino-Bodas et al. 2010a, 2013b). However, in the present study the sampling is expanded substantially, including specimens from numerous countries and species not sampled to date, or poorly studied, such as C. macrophyllodes, C. trassii or C. stricta. The phylogenetic placement of the species of the C. cervicornis group is consistent with the results obtained by Stenroos et al. (2019), with all of them belonging to the Cladonia subclade.

The phylogenetic relationships of species in the C. cervicornis group, within the Cladonia subclade, are uncertain. Our results indicate that C. verticillata s. str. is related to C. subulata, a sorediate species without central proliferations, although this relationship was only supported in the Bayesian analysis (Fig. 1). This is a similar result to that found by Beiggi & Piercey-Normore (2007). However, the phylogeny of Stenroos et al. (2019) showed a possible relationship between C. verticillata and C. pulvinata, although it was not supported.

In our phylogeny, some clades consisted exclusively of species with central proliferations, but these clades did not always include species with morphological or biogeographical similarities. An example is the clade formed by C. andesita, C. centrophora, C. ceratophyllina, C. nitens and C. teuvoana. While C. teuvoana is restricted to Europe, the rest of the species have very different distribution patterns. Cladonia ceratophyllina and C. centrophora are two morphologically similar African species (Swinscow & Krog 1988; Ahti & Aptroot 1992). Cladonia andesita was originally described from Colombia, but other populations later found in Africa, with some morphological differences, could represent a different species (Swinscow & Krog 1988; Ahti 2000). Cladonia nitens has an amphi-Beringian distribution and is characterized by having slender, shiny podetia and a large melanotic zone (Ahti 2007).

Another clade, consisting exclusively of species with central proliferations, is formed by C. aleuropoda Vain., C. calycantha Delise ex Nyl., C. cineracea, C. isabellina Vain., C. mawsonii C.W. Dodge, C. melanopoda Ahti and a specimen of C. rappii. This clade was previously shown in the phylogeny of Stenroos et al. (2019) but without support. With the exceptions of C. mawsonii, which has a disjunct distribution between Australasia and Tierra del Fuego, and C. rappii, with a broader distribution (Stenroos & Ahti 1990; Ahti 2000), the remaining taxa in the clade are neotropical, abundant in the páramos of the Andes and morphologically very similar. The delimitation of neotropical species is still not well understood and needs further study (Ahti 2000).

The phylogenetic relationships of C. trassii are studied here for the first time. Although numerous specimens were selected for phylogenetic analysis, sequencing failed for most of the gene regions (Table 1). Only for IGS rDNA were a high number of sequences obtained. Phylogenetic analyses of this region revealed that the specimens identified as C. trassii formed a polyphyletic group. A similar result was found in the multilocus analysis, which included only those specimens with sequences for at least two loci. This result is not surprising, since C. trassii is an extremely variable species, difficult to distinguish from C. stricta and C. uliginosa (Ahti 1998; Ahti & Stenroos 2013). However, phylogenetic analyses did not show that C. trassii specimens are related to C. uliginosa. Our results point out a possible relationship between C. uliginosa and C. macrophyllodes, although it was supported only in the Bayesian analysis. These two species are easily distinguishable. Cladonia macrophyllodes has a well-developed and persistent primary thallus, while C. uliginosa has an inconspicuous and evanescent primary thallus. The two species also show considerable discrepancies in the morphology of the podetia. Cladonia uliginosa has podetia with a discontinuous cortex, with several tiers of proliferations and a strongly melanotic base. None of these characters are shared by C. macrophyllodes, which has podetia with a continuous cortex and abundant squamules on the scyphi, and tiers are rarely present. However, both species synthesize the same secondary metabolites: fumarprotocetraric acid and atranorin (Ahti 1978; Hansen & Ahti 2011; Ahti & Stenroos 2013).

Our analyses did not resolve the phylogenetic relationships of C. stricta. IGS rDNA analysis showed that the specimens of this species formed a polyphyletic group (based on three sequences) and the only RPB2 sequence obtained for it showed a relationship with a specimen of C. trassii. However, these data are not enough to draw conclusions about the taxonomy and phylogenetic relationships of this species. Based on its morphological features, C. stricta has been related to C. phyllophora Hoffm., a species characterized by podetia with a strongly melanotic base, with lateral proliferations that open and rarely produce central proliferations (Osyczka 2006; Hansen & Ahti 2011; Ahti & Stenroos 2013). Our phylogenetic analyses included several specimens of C. phyllophora, which was resolved as a monophyletic group, confirming the findings of Fontaine et al. (2010). However, C. phyllophora does not belong to the subclade Cladonia, as the phylogeny of Stenroos et al. (2019) showed, but belongs to the subclade Foliosae.

The phylogeny of the subclade Cladonia is in general still poorly resolved (Fig. 1). Although some well-supported clades have been found within this subclade, the relationships among them are not supported. This result highlights the need for future studies, including sampling in other geographical regions, that will help to better understand both the phylogenetic relationships of species with central proliferations as well as species delimitation. Preliminary data for some species such as C. rappii reveal that they may include more than one species and therefore a thorough morphological and molecular study is necessary to clarify their taxonomy. The use of phylogenomic techniques that increase the number of loci will allow better-founded phylogenetic hypotheses to be established.

Taxonomy

Cladonia teuvoana Pino-Bodas, Burgaz & Aptroot

MycoBank No.: MB 855855

Primary thallus persistent, with small squamules, white on the lower surface. Podetia with narrow scyphi, abruptly widening, with a corticate surface.

Type: The Netherlands, Drenthe, Gasterense zand, on sand, 24 March 2018, A. Aptroot 76154 (MACB 127525—holotype). GenBank Accession nos.: PQ252415, PQ252566, PQ254487.

(Fig. 4A–C)

Primary thallus persistent, prostrate or ascendent, squamules 3.7–8.11(9.2) mm long × 3.4–5.8 mm wide, apices sometimes incurved, green to greenish brown on the upper side, white on the lower side, darkening towards the base (up to 1/3 of squamule length). Podetia 4.7–8.7 mm tall, slender, simple, scyphose, rarely with squamules at the base, brownish, without a melanotic base; scyphi regular, with margins entire, very narrow, 1–3 mm wide, abruptly tapering; surface corticate, cortex smooth in the upper part, verrucose in old parts, faintly pruinose towards the scyphi; podetial wall 183–223 μm wide. Hymenial discs infrequent, dark brown, no well-developed ones observed; ascospores not observed.

Conidiomata at margin of scyphi, globose, pycnidial slime hyaline; conidia 7–10 × 1 mm, hyaline, falciform.

Chemistry

PD+ red, K−, UV−; contains the fumarprotocetraric acid complex.

Etymology

This species is named in honour of Prof. Teuvo Ahti, on the occasion of his 90th birthday, for his great contribution to lichenology.

Habitat and distribution

Cladonia teuvoana occurs on heathlands and sandy areas, and is currently known from Denmark, the Czech Republic, the Netherlands, Spain, Sweden and the United Kingdom.

Remarks

The species differs from C. cervicornis s. str. in having a less well-developed primary thallus with a white underside and narrower scyphi that abruptly widen. It can be distinguished from C. pulvinata by the absence of psoromic acid and differs from C. verticillata in having narrow scyphi and a smaller primary thallus. Cladonia teuvoana probably has a wider distribution in Europe. To achieve a more realistic understanding of its distribution, it would be necessary to conduct a comprehensive study of herbarium material.

Additional specimens

Great Britain: England: V.C. 11, South Hampshire, New Forest National Park, Lyndhurst, Bolton Bench, heather with Calluna vulgaris and Erica cinerea, 50°52ʹ17.2ʺN, 1°33ʹ23ʺW, alt. 36 m, 2019, R. Pino-Bodas & N. Sanderson (MACB).—Denmark: Jutland, Harrild, Hede, 62.07550°N 5.10650°E, 2016, U. Søchting 12520 (C-L 75831).—The Netherlands: Drenthe: Balloo, Ballooerveld, on sand, 2024, Aptroot 76155 (MACB).—Czech Republic: C. Bohemia: Praha, Natural Park Košíře-Motol, path of a heath with grove of Quercus petraea, Betula pendula and Pinus sylvestris with Calluna vulgaris in understorey, 50°03ʹ45.6ʺN, 14°19ʹ48.7ʺE, alt. 338 m, 2018, Z. Palice & P. Uhlík 24917.—Spain: Guipúzcoa: Oñate, Aizkorri-Aratz Natural Park, Malkorra, heathland with sandstone, 42°56ʹ36.5ʺN, 2°21ʹ51.9ʺW, alt. 1500 m, R. Pino-Bodas 254-2023 (MACB 125083).—Sweden: 2005, B. P. Løfall 10970 (H).