We analyzed a total of 339 specimens (Supplementary Table 1) belonging to three closely related fungal genera Cladonia (n=179), Lepraria (n=23), and Stereocaulon (n=137) collected between years 2011 and 2020 on the Canary Islands (El Hierro, La Palma, La Gomera, Tenerife, Gran Canaria, Lanzarote), Madeira, Sicily, and the Aeolian Islands (Vulcano, Salina, Stromboli; Supplementary Figure 1). The sampling sites represented diverse habitats from malpaíses (i.e., barren landscapes) to a laurel forest, various types of substrates, and were located at a range of altitudes 75–2,360m. We identified the mycobiont morphospecies using standard morphological methods, taxonomical keys (e.g., Lamb, 1978; Baruffo et al., 2006; Burgaz and Ahti, 2009) and thin-layer chromatography (TLC) on Merck silica gel 60 F254 pre-coated glass plates in solvent systems A, B, and C, according to Orange et al. (2001). For the purpose of describing Stereocaulon species, measurements and observations were done in water, except for ascospores and paraphyses observed in KOH. Lichen specimens were deposited in the Herbarium of the Institute of Botany, the Czech Academy of Sciences (PRA), and the private herbarium of J. Malíček.

We extracted DNA directly from lichen thalli (total lichen DNA) following the standard CTAB protocol for lichens (Cubero et al., 1999). Firstly, we examined thalli under a dissecting microscope and washed them with running water (except for Lepraria thalli) before DNA extraction to prevent contamination by soredia from other lichens. We amplified the algal and fungal nuclear internal transcribed spacer (ITS, ITS1-5.8S-ITS2 rDNA) and the algal actin type I gene (including one complete exon and two introns located at codon positions 206 and 248; Weber and Kabsch, 1994) using primers listed in Table 1. The PCR conditions were as described in Vančurová et al. (2018). Every PCR run included negative controls, without a DNA template. The PCR products were purified using by NucleoMag® NGS Clean-up and Size Select kit (Macherey-Nagel, Duren, Germany) and sequenced using the same primers at Macrogen in Amsterdam, Netherlands. The 507 newly obtained sequences of the ITS rDNA and actin type I regions are available in GenBank under accession numbers OL622077–OL622095 and OL625120–OL625607 (Supplementary Table 1).

We prepared individual sequence alignments for Stereocaulon, Cladonia, Lepraria, Chloroidium, Vulcanochloris (ITS rDNA), and Asterochloris (ITS rDNA and actin type I) datasets using MAFFT v.7 software (Katoh and Standley, 2013). The sequences obtained for Asterochloris were analyzed as a single locus dataset for the ITS rDNA (data not shown) and as a concatenated dataset of ITS rDNA and actin type I loci. In the case of Cladonia, we used Gblocks to remove introns from alignment and to eliminate poorly aligned positions (Castresana, 2000).

Supplementary Table 2 summarizes a total number of sequences included in particular alignments, newly obtained sequences, previously published sequences originated from the study area, reference sequences retrieved from GenBank, and unique sequences after deleting identical ones. All DNA alignments are freely available on Mendeley Data: http://dx.doi.org/10.17632/428v52svtp.1.

Phylogenetic trees were inferred with Bayesian Inference (BI) using MrBayes v.3.2.7a (Ronquist and Huelsenbeck, 2003; Ronquist et al., 2012), maximum likelihood (ML) analysis using GARLI v.2.0 (Zwickl, 2006), and maximum parsimony (MP) analysis using PAUP v.4.0b10 (Swofford, 2003). BI and ML analyses were carried out on a partitioned dataset to differentiate among ITS1, 5.8 S and ITS2 rDNA, actin intron 206, actin intron 248, and actin exon regions. Substitution models (Supplementary Table 3) were selected using the Bayesian information criterion (BIC) as implemented in JModelTest2 (Guindon and Gascuel, 2003; Darriba et al., 2012). Two parallel MCMC runs, with four chains, were carried out for 10 and 5 million generations for Asterochloris and other datasets, respectively. Trees and parameters were sampled every 100 generations. Convergence of the two cold chains was assessed during the run by calculating the average SD of split frequencies (SDSF). The SDSF value between simultaneous runs was < 0.01 in all cases. Finally, the burn-in values were determined using the “sump” command. ML analysis was carried out using default settings, five search replicates, and the automatic termination set at 5 million generations. The MP analysis was performed using heuristic searches with 1,000 random sequence addition replicates and random addition of sequences (the number was limited to 10⁴ per replicate). ML and MP bootstrap support values were obtained from 100 to 1,000 bootstrap replicates, respectively. Only one search replicate was applied for ML bootstrapping.

We maintained the nomenclature and species boundaries delimited in recent studies (Škaloud et al., 2015; Vančurová et al., 2018, 2020; Kosecka et al., 2021) for Asterochloris species and the species boundaries delimited by Vančurová et al. (2018) and the nomenclature used therein (OTU1–OTU57) for Stereocaulon species-level lineages.

We depicted the interactions between phycobiont species-level lineages in the genus Asterochloris and mycobiont genera and between phycobiont species-level lineages and mycobiont species-level lineages in the genus Stereocaulon as interaction networks produced using bipartite package (Dormann et al., 2008) in R.

The climatic niche of the most abundant species-level lineages of phycobionts and Stereocaulon mycobionts were represented using the Hutchinsonian niche concept that describes a species niche as an n-dimensional hypervolume, where the dimensions are environmental variables (Hutchinson, 1957). In the present study, these environmental dimensions were defined based on 19 Bioclim variables (Karger et al., 2017). We constructed the climatic hypervolumes by multivariate kernel density estimation (Blonder et al., 2014). First, we performed the PCA analysis of 19 Bioclim variables to reduce the total number of predictors. The first two PCA axes (explaining 82% of the total variance) were then selected to calculate hypervolumes for each species-level lineage and genus. The boundaries of the kernel density estimates were delineated by the probability threshold, using the 0.85 quantile value. To project the niche spaces of particular lineages, hypervolume contours were plotted based on 5,000 random background points, using the alphahull contour type and alpha smoothing value 0.55. The analyses were performed in R, using the hypervolume (Blonder et al., 2014) and alphahull (Pateiro-Lopez and Rodriguez-Casal, 2016) packages.

To assess the influence of climatic variables on phycobiont distribution, the contrasting phycobiont groups were compared for each Bioclim variable and altitude using Welch’s two-sample t-test. We compared: (1) Asterochloris mediterranea (as the most abundant phycobiont in the study area) with other Asterochloris spp. in association with all mycobionts, (2) A. mediterranea with other Asterochloris spp. in association with Cladonia rangiformis (as the most abundant Cladonia species in the study area) mycobiont, (3) A. mediterranea with other Asterochloris spp. in association with Stereocaulon azoreum mycobiont (as the most abundant Stereocaulon species in the study area associated with Asterochloris spp.), and (4) Chloroidium ellipsoideum with C. lichenum A (sensu Darienko et al., 2018) in association with Stereocaulon vesuvianum (OTU11) mycobiont (as the most abundant Stereocaulon species in the study area). We tested phycobionts in pairs with their mycobionts to inspect ecological requirements of the symbiotic system (unlike to separate analyses of symbiotic partners in previous studies, e.g., Vančurová et al., 2018). We could not compare those groups of phycobionts associated with Lepraria spp. mycobionts due to the low number of samples. We performed the analyses in R using function t.test.

We recovered Stereocaulon mycobionts from the study area in six lineages (Figure 1): OTU13 (identified morphologically as S. azoreum and clustered with S. azoreum sequences), OTU11 (identified morphologically as S. vesuvianum), OTU3 (identified morphologically as S. pileatum and clustered with S. pileatum and S. octomerellum), OTU52 (described here as a new species Stereocaulon canariense; sister to S. virgatum sequence DQ396964), OTU22 (clustered with S. delisei and S. corticatulum), and OTU23 (clustered with S. atlanticum and S. ramulosum). The most frequent species in the study area was S. vesuvianum (47% of the Stereocaulon samples).

A Bayesian analysis of the ITS rDNA of Cladonia mycobionts resulted in 22 supported Cladonia lineages (Supplementary Figure 2). The most recovered Cladonia species was C. rangiformis (30% of the Cladonia samples), followed by C. humilis (20% of the Cladonia samples). In most cases, phylogenetically separated lineages corresponded to morphologically delimited species; however, the lineage “Cladonia sp. 1” contained several morphologically different entities. It has been previously reported that ITS rDNA may fail to distinguish Cladonia species in some cases (Kelly et al., 2011; Kanz et al., 2015); however, it has been suggested as a Cladonia barcode marker at the same time (Pino-Bodas et al., 2013).

The phylogenetic hypothesis resulting from the Bayesian analysis of the ITS rDNA sequences of Lepraria (Supplementary Figure 3) was congruent with that of previous studies (Crespo et al., 2006; Tretiach et al., 2009). We recovered Lepraria mycobionts from the study area in five lineages. The most abundant was L. santosii (55% of Lepraria samples), originally described from Tenerife (Crespo et al., 2006).

A phylogram resulting from the Bayesian analysis of ITS rDNA and actin type I sequences of Asterochloris is shown in Figure 2. We recorded phycobionts from the study area in eight species-level lineages. The overwhelming majority of samples belonged to A. mediterranea (72% of Asterochloris samples). A. woessiae and A. italiana were also relatively common (16 and 7%, respectively). Other lineages were rather rare. The lineage StA10 is a new, highly resolved lineage in Asterochloris.

The phylogenetic hypothesis resulting from the Bayesian analysis of the ITS rDNA sequences of Chloroidium (Figure 3) was congruent with that of previous studies (Darienko et al., 2018; Vančurová et al., 2018). Since C. lichenum became a paraphyletic species (divided into three clades: A, B, and C) after the last taxonomic revision (Darienko et al., 2018), we considered the two monophyletic clades C. lichenum A and C. lichenum B as species level-lineages (following species delimitation analyses; Vančurová et al., 2018). We recovered Chloroidium phycobionts from the study area in three species-level lineages: C. ellipsoideum (45% of Chloroidium samples), C. lichenum A (50% of Chloroidium samples), and C. lichenum B.

Furthermore, we recovered Vulcanochloris phycobionts from the study area in three lineages corresponding to three species (Figure 4): V. symbiotica (70% of Vulcanochloris samples), V. guanchorum, and V. canariensis.

Finally, we confirmed the identity of sample A194 by a BLAST search against the GenBank database. The most similar hit (95% sequence identity) was Myrmecia israliensis KY981668. Therefore, we labelled this sample as Myrmecia sp.

In total, we found 15 species-level lineages within four genera of phycobionts, from one to nine on each island (Figure 5). Madeira was identified as the most phycobiont species-rich island in our study area. The number of species-level lineages was highest (4.1) also after rarefaction to seven samples, which was the smallest number of samples per island in the dataset.

Asterochloris mediterranea represented the most abundant phycobiont in our dataset, was present on all studied islands and formed associations with all three mycobiont genera (Figures 5, 6A). However, its prevalence in the thalli of different genera of mycobionts varied. Asterochloris mediterranea was present in 83, 73, and 28% of the thalli involving Asterochloris phycobionts in association with Cladonia, Lepraria, and Stereocaulon, respectively (Figure 6B). Asterochloris woessiae and Asterochloris StA10 were also shared by all three mycobiont genera and present on multiple islands. On the other hand, A. italiana and A. glomerata associated solely with representatives of Stereocaulon and Cladonia.

Chloroidium ellipsoideum associated with S. vesuvianum and Stereocaulon OTU23 (Figures 5, 6B), was the second most geographically widespread phycobiont, and was present on eight islands (absent on Gran Canaria, Lanzarote, and Isola Salina).

The phycobionts Myrmecia sp. (associated with a Cladonia mycobiont), Asterochloris clade A11 (with Cladonia), and A. stereocaulonicola (with Stereocaulon) were recorded only once in our dataset. The phycobiont A. aff. italiana was found in three samples but only on a single island (Madeira) and was associated with a single mycobiont (Stereocaulon azoreum).

Vulcanochloris phycobionts were restricted to the Canary Islands, namely La Palma, Lanzarote, El Hierro, and Gran Canaria. On La Palma and Lanzarote, all three Vulcanochloris species were present. Interestingly, Vulcanochloris phycobionts were found exclusively in association with S. canariense (Figure 6B).

We constructed two-dimensional (PC1-PC2, explaining 82.4% of the variation in climatic variables) hypervolumes for the six most abundant algal species-level lineages and the three most abundant Stereocaulon mycobiont species-level lineages. Since the climatic characteristics differed considerably between Macaronesia and the Mediterranean region, the hypervolumes split into two isolated parts. Within each of them, there was a gradient of climatic niches along PC1, representing mostly the gradients of temperature and precipitation (Figure 7).

Among the algal species-level lineages, Vulcanochloris symbiotica tolerated the driest and warmest climate. On the other hand, Asterochloris woessiae, A. italiana, and Chloroidium lichenum A were distributed predominantly in the more humid and relatively colder areas. Asterochloris mediterranea and C. ellipsoideum were widely distributed in warm and dry areas and their climatic niches partly overlapped with that of both sides of the continuum. Climatic niches of A. mediterranea, C. ellipsoideum, and A. woessiae involved the whole range of temperature and precipitation seasonality (corresponding roughly with PC2). Remarkably, the niches of A. mediterranea and C. ellipsoideum overlapped to a large extent. Similarly, the niches of A. italiana and C. lichenum A were equivalent.

The climatic hypervolume of S. vesuvianum was the widest among Stereocaulon species. It included the whole range of temperature seasonality in Macaronesia as well as that of the Mediterranean region. The warmest and driest areas with high precipitation seasonality were dominated by S. canariense. The climatic niche of S. azoreum overlapped with part of that of S. vesuvianum. These two species likely differed in microclimatic factors (not included in this study).

Finally, we focused on the niche of the most abundant phycobiont in the study area, A. mediterranea. We inspected the differences between the climatic requirements of this species and those of other Asterochloris species associated with either two dominant mycobiont species (Cladonia rangiformis and Stereocaulon azoreum) or all mycobionts using Welch’s t-tests. Similarly, we tested the differences between Chloroidium ellipsoideum and C. lichenum A. Table 2 summarizes the results of the tests performed for various combinations of phycobionts and mycobionts. The means of three variables showed significant differences (α=0.01) across all combinations: BIO1 – annual mean temperature, BIO8 – mean temperature of the wettest quarter, and BIO11 – mean temperature of the coldest quarter (Figure 8). In contrast, we did not find any significant difference between the groups in BIO3 [Isothermality (BIO2/BIO7)] and BIO4 (Temperature Seasonality) climatic variables.

Interestingly, along the gradient of annual mean temperature, A. mediterranea was replaced by other Asterochloris species at different temperature levels depending on whether it was associated with Stereocaulon azoreum or C. rangiformis (the later in about 1°C colder temperature). Furthermore, C. ellipsoideum was replaced by C. lichenum A at an annual mean temperature of approximately 13°C (Figure 8).

Stereocaulon canariense Malíček & Vančurová sp. nov.

Mycobank: MB841859.

We found 15 species-level lineages of phycobionts in 338 lichen thalli belonging to the genera Stereocaulon, Cladonia, and Lepraria. Out of these 15 lineages, only eight belonged to the genus Asterochloris. This is a relatively surprising finding considering the results of previous studies focusing on photobiont diversity of corresponding mycobiont genera. For example, Kosecka et al. (2021) found 21 Asterochloris and one Vulcanochloris species-level lineages in Bolivia with a considerably smaller sampling of the same lichen genera. Vančurová et al. (2020) sampled only two Stereocaulon species-level lineages in the Swiss Alps and found 14 Asterochloris and three other trebouxiophycean species-level lineages. Peksa and Škaloud (2011) recovered 10 Asterochloris lineages associated with Lepraria in the Czech Republic. In the study area, globally abundant Asterochloris species (Asterochloris irregularis, A. pseudirregularis, A. lobophora, A. phycobiontica, A. gaertneri, A. friedlii, and others) were not detected. One of the most common Asterochloris species, A. glomerata (Pino-Bodas and Stenroos, 2020), was present in three samples and belonged to the single genotype that is known to be exclusively from the study area.

We hypothesized that the isolation of oceanic islands may explain the limited phycobiont species richness. However, the islands in Macaronesia showed higher species richness than the less isolated Sicily and Aeolian Islands. It is more likely that phycobiont distribution can be restricted by a limited number of substrate/habitat types. The main reason for the absence of many Asterochloris lineages might be their preference for cold and wet climates (Vančurová et al., 2018; Pino-Bodas and Stenroos, 2020). Nevertheless, there were a few cases in which phycobionts that are typical of cold climates were recorded; for example, one sample included A. stereocaulonicola, which was originally described in Antarctica (Kim et al., 2020). In contrast, A. mediterranea, the most abundant phycobiont in the study area, copes well with warm and dry climates (Moya et al., 2015; Pino-Bodas and Stenroos, 2020). Madeira, the wettest and coldest island in our study, showed the highest phycobiont species richness. Interestingly, we found all three Vulcanochloris species with a high haplotype diversity on the warmest and driest island in the study area, Lanzarote. This island seemed to be the hot-spot of Vulcanochloris diversity. In contrast, lichens associated with Asterochloris (including A. mediterranea, which performs well in warmer habitats) were extremely rare in that island. Lindgren et al. (2020) proposed a hypothesis explaining the association between Haveochlorella and several Sticta spp. on Madagascar and Reunion. A long-distance dispersal of mycobionts may have been followed by an association with locally adapted phycobionts and the subsequent diversification of mycobionts. This pattern can be true also for the relationship between S. canariense and Vulcanochloris spp.

In one case, we noticed Myrmecia sp. as the phycobiont of Cladonia. The genus Asterochloris has generally been considered the exclusive phycobiont of this broadly distributed mycobiont genus (Piercey-Normore and DePriest, 2001; Yahr et al., 2006; Steinová et al., 2019; Pino-Bodas and Stenroos, 2020). However, its ability to associate with Trebouxia and Chloroidium under stressful conditions has recently been demonstrated (Park et al., 2015; Osyczka et al., 2020). Moreover, Elshobary et al. (2015) discovered unknown trebouxiophycean algae as phycobionts of Cladonia macrophylla.

Kaasalainen et al. (2021) demonstrated that cyanolichens share photobionts and revealed the presence of photobiont-mediated lichen guilds. Most of the mycobionts shared photobionts with other non-related fungal species but other photobiont-mycobiont pairs remained isolated. In our study area, the vast majority of Cladonia and Lepraria mycobionts associated with the A. mediterranea phycobiont. On the other hand, phycobionts of Stereocaulon were highly diverse. After downsampling to a sample size of 22 (the number of Lepraria samples) Stereocaulon mycobionts associated with almost two times more phycobiont species than Cladonia or Lepraria (Supplementary Figure 5). The Stereocaulon species-level lineages showed a high level of specificity towards their phycobionts (on the level of algal genera, but not on the level of algal species). Stereocaulon azoreum was associated with the same phycobionts as Cladonia and Lepraria on the Canary Islands, whereas in Madeira, it associated with another algal species. Stereocaulon vesuvianum and S. canariense were associated with Chloroidium spp. and Vulcanochloris spp., respectively. Although Cladonia and Lepraria mycobionts often grow together with Stereocaulon mycobionts, S. vesuvianum and S. canariense never shared phycobionts with them in our samples. Moreover, at localities where more than one thallus was sampled with Asterochloris, the number of species-level lineages varied. In some cases, all samples shared one species, but at other localities, each sample contained different species of Asterochloris.

Stereocaulon is not the only mycobiont genus able to associate with various phycobiont genera. The mycobiont genus Sticta has been shown to associate (besides cyanobacteria) with multiple trebouxiophycean algae (Haveochlorella, Chloroidium, Symbiochloris, and Elliptochloris); some Sticta species are specific towards particular phycobionts, but others can associate with up to three trebouxiophycean genera (Lindgren et al., 2020). Previous studies hypothesized that Chloroidium phycobionts are well-adapted to volcanic substrates (Vančurová et al., 2018). However, Cladonia mycobionts in association with Asterochloris can also cope with this substrate. This fact points to the role of specificity or other mycobiont features on phycobiont distribution. On the other hand, S. canariense was able to grow at arid localities in association with Vulcanochloris spp. Nevertheless, we could not find any lichen usually associated with Asterochloris (Cladonia and Lepraria) in the most arid parts of the study area (Fuerteventura and most of the localities on Lanzarote), even though these have previously been reported (van den Boom and Etayo, 2006). In contrast, the driest areas were dominated by lichens known to be associated with Trebouxia (for example, Ramalina spp.). The ability of some Cladonia species to associate with Trebouxia phycobionts (Osyczka et al., 2020; Shishido et al., 2021) might be helpful under such conditions (Romeike et al., 2002; de Vera, 2012; Sadowsky and Ott, 2012; Candotto-Carniel et al., 2016; Leavitt et al., 2016); however, we did not find any sample containing Trebouxia.

In a few cases, we found Asterochloris phycobionts in thalli of S. vesuvianum and S. canariense, even though these have high specificity towards other phycobiont genera. In some of these cases, we also found common phycobionts in the same thallus. In the case of S. vesuvianum, we repeatedly detected more than one Chloroidium species in a single thallus. Furthermore, we were almost unable to obtain a non-mixed sequence of phycobionts of S. vesuvianum at the localities on the slopes of Mt. Etna. In that area, Darienko et al. (2018) reported multiple free-living Chloroidium species that can coexist in Stereocaulon thalli. Paul et al. (2018) mentioned difficulties with Sanger sequencing as a possible indicator of algal plurality (i.e., the co-occurrence of multiple phycobionts in individual lichen thalli). Several authors found phycobionts with different ecological optima in a single thallus (Casano et al., 2011; Gasulla et al., 2020; Molins et al., 2020), which was also the case of multiple Chloroidium spp. in S. vesuvianum thalli. Vančurová et al. (2020) found additional algal phycobionts in Stereocaulon thalli, which were otherwise detected as the predominant phycobionts in other mycobiont species at the same locality. Therefore, we believe that phycobiont sharing is possible even in cases when the lichens differ in a predominant phycobiont. Since different lichen genera and species differ in the incidence of this phenomenon (Dal Grande et al., 2017; Smith et al., 2020), we cannot reliably estimate whether other lichen genera in the study area also show such algal plurality.

Even though almost all phycobionts in the study area are known to be well adapted to warm areas, they spread out along the climatic gradient. We examined the differences between those phycobionts distributed in the relatively colder and more humid areas and those taking up the widest space in the middle of the climatic niche (Figure 7B). Since both mycobionts and phycobionts contribute to the ecology of the holobiont, we tested various combinations of partners. Asterochloris mediterranea and other Asterochloris phycobionts in Stereocaulon azoreum differed significantly in 17 out of 19 tested climatic variables. However, this contrasting pair of phycobionts in thalli of C. rangiformis showed significant differences along four variables related to temperature. We considered the possibility that different thallus structure of lichen genera affected their water regime as suggested by Sadowsky et al. (2012); however, the pair of Chloroidium species in thalli of S. vesuvianum also differed somewhat in temperature-related variables. Stereocaulon azoreum was distributed only in Macaronesia, where temperature and precipitation are correlated. Hence, it is impossible to estimate what factor is crucial for driving the distribution of their phycobionts in this area. On the other hand, S. vesuvianum and C. rangiformis both grew on Sicily and the Aeolian Islands, where the climate characteristics are different. The three factors that were significant in all combinations were related to temperature, emphasizing the importance of this factor in the selection of the appropriate phycobiont. Several previous studies also found temperature to be the critical factor determining the distribution of phycobionts (Vančurová et al., 2018; Molins et al., 2020; Rolshausen et al., 2020). Rolshausen et al. (2020) recently proposed parallel symbiont turnover zones as demarcated regions where symbiont replacement is most likely to occur. In all gradients, this symbiont turnover zone is characterized by approximately 12°C average annual temperature. We found this pattern in all combinations, but the average annual temperature of the turnover zone was slightly higher and not uniform among the mycobiont species/genera (Figure 10). We attribute this difference to the additional influence of mycobiont selectivity. Previous studies did not determine specific turnover zones, but described the gradual change of phycobionts along the altitudinal gradient (Vargas Castillo and Beck, 2012; Gasulla et al., 2020). Moreover, Molins et al. (2020) studied Trebouxia phycobionts of Buellia zoharyi on Lanzarote, Fuerteventura, and Tenerife. They found three species that differed on their distribution ranges and the level of their tolerance to high temperatures (even under laboratory conditions). Interestingly, in corals, which is another symbiotic association similar to lichens in many aspects, the ability to cooperate with symbionts characterized by various temperature optima has been documented as an advantageous adaptive strategy (Silverstein et al., 2015). Similarly, under a scenario of global change, it may also be a critical factor for lichens (Ellis, 2019), favoring those with the ability to associate with several phycobionts covering a greater temperature gradient. However, it is possible that Cladonia, Lepraria, and Stereocaulon mycobionts or their phycobionts reached their maximum temperature limits, thus restricting their distribution ranges, because we did not find them on the warmest sites of the study area (Chazarra et al., 2011).

Macaronesia is known for high level of endemism (Fernández-Palacios and Whittaker, 2008). As an oceanic island system, it is an attractive object for evolution studies (Emerson, 2002). Besides well-known examples of beetles and angiosperms, examples of lichen speciation have been documented within this area (Sérusiaux et al., 2011). Stereocaulon canariense was recorded only in the Canary Islands so far. However, its phycobiont, Vulcanochloris, has been rarely reported from other parts of the world as phycobiont of Stereocaulon (Kosecka et al., 2021) or other lichens (Vaiglová, 2017). The communication between Macaronesian and Mediterranean biota has been repeatedly documented (Carine et al., 2004; Vondrák et al., 2020). Consequently, other localities of this species can be found in the future. Moreover, the origin of this possibly endemic lichen can be a topic of future research.