Together Apart: Evaluating Lichen-Phorophyte Specificity in the Canarian Laurel Forest

The effects of host tree identity on epiphyte lichen communities are a controversial issue, as the results obtained in different forest environments studied are not consistent. We investigated the host preferences for lichens in the laurel forest of Macaronesia. For this purpose, we analyzed the lichen communities growing on the four most common trees (Erica canariensis Rivas-Mart., M. Osorio and Wildpret, Morella faya (Aiton) Wilbur, Laurus novoca-nariensis Rivas-Mart., Lousa, Fern. Prieto, E. Días, J.C. Costa and C. Aguiar, and Ilex canariensis Poir. in Lamarck) in the laurel forest of the Canary Islands. The diversity, richness, and lichen composition showed a repetitive and common pattern with the functional traits studied. Although the existence of specificity with respect to the phorophyte species was not demonstrated, there was a clear affinity of the epiphytic lichens to the physico-chemical features of the bark (texture and pH), canopy architecture, foliar characteristics, etc. Our results highlight the importance of the natural diversity of tree species in the laurel forest. Due to the diversity and uniqueness of the lichen species that support each of the phorophytes, this fact should be taken into account in landscape protection and restoration actions, especially in those islands where the forest is highly fragmented.


Introduction
The Macaronesian laurel forests, also called laurisilva and monteverde, are humid to hyper humid evergreen forests of the cloud belt of the Macaronesian islands. They are a value not only in terms of biodiversity but also biogeography [1]. Tree species with laurel-shaped leaves are predominant, forming a dense canopy up to 40 m high with scant vegetation in the understory [2]. Laurel forest is cataloged as a priority habitat by the European Union [3], and it is an example of a biodiversity hotspot threatened by human impact. Mankind has been altering the native forests in Macaronesia for at least five centuries. In the Canary Islands, these disturbance activities began with the aboriginal inhabitants; although they did not have sophisticated tools, they certainly used fire to create farmland and introduced livestock (sheep, goats, and other animals), as well as other plants [4,5]. After the Hispanic colonization in the 15th century, the natural landscape became increasingly transformed. In the second half of the 20th century, a significant part of this native laurel forest was cleared in the Canaries for livestock, agriculture, and planting exotic species. Due to this extensive exploitation, only 11.8% of the original laurel forest remains in this archipelago nowadays [5].
Cloud forests are considered one of the most vulnerable terrestrial ecosystems to climate change [6]. Global climate models predict a future downward shift of the cloud base in the Macaronesian region [7]. Therefore, climate change would presumably expose the laurel forests to higher radiation and temperatures and lower rainfall [8]. These changes in climatic conditions could produce an extensive change in a plurispecific forest, such as Monteverde, which has more than 20 different species in its tree stratum. Among these This study was carried out in the Canary Islands. The main areas where the laurel forests are growing and are more abundant are located in Tenerife, La Gomera, and La Palma, so we restricted the fieldwork to these islands. The other islands do not have a proper representation of this forest because: (i) the island does not have the environmental conditions required for its presence, or (ii) logging over the last five centuries has eliminated this endemic forest [36][37][38].
Macaronesian cloud forests comprise an environmentally complex system in which different plant associations of evergreen laurel forest can be distinguished. While macroclimatic conditions in all these stands are quite similar, there are small differences in some abiotic variables, such as the altitude, exposure, topography, fog regimes, etc., so microclimatic conditions may be different and could affect the lichen communities. To minimize the effect of these differences and to homogenize our plots as much as possible, all of them were always located in the potential habitat of the humid evergreen laurel forest (Lauro novocanariensis-Perseetum indicae Oberdorfer ex Rivas-Martínez, Arnáiz, Barreno andA. Crespo 1977 corr. Rivas-Martínez et al., 2002), the most widespread association in the archipelago following Ref. [5]. (Figure 1). The best examples of this forest are found in the north of Tenerife (N), E of La Palma, and on La Gomera [39]. The fieldwork was carried out on these islands between 750 and 1100 m a.s.l. where the humid laurel forest grows ( Table 1).
The temperatures in these areas are mild throughout the year, with no frost or snow in winter. The mean annual temperature varies from 13 to 18 • C. Annual rainfall ranges from 500 to 1200 mm with a dry period in summer [40]. However, the study area is affected by the prevailing NE trade winds, which are responsible for the frequent low-level cloud

Plot Selection and Sampling
Our field campaigns were authorized by Área de Medioambiente-Cabildo Insular de Tenerife, by Área de Medioambiente-Cabildo Insular de La Palma, and by Parque Nacional de Garajonay, with permission to have access and collect data (including biological material).
Six sampling plots of 10 × 10 m were chosen on each island in well-preserved areas without signs of anthropogenic disturbances. In each plot, we selected two trees of Morella

Plot Selection and Sampling
Our field campaigns were authorized by Área de Medioambiente-Cabildo Insular de Tenerife, by Área de Medioambiente-Cabildo Insular de La Palma, and by Parque Nacional de Garajonay, with permission to have access and collect data (including biological material).
Six sampling plots of 10 × 10 m were chosen on each island in well-preserved areas without signs of anthropogenic disturbances. In each plot, we selected two trees of Morella faya, two of Erica canariensis, two of Laurus novocanariensis, and two of Ilex canariensis ( Figure 2). The laurel forest is a plurispecific forest, with about twenty tree species. Studies on host tree specificity in natural forests, where the scarcity of individuals of a given group for the many coexisting species prevents adequate replication, have often yielded inconclusive results in some groups of epiphytes [42]. Host tree species can support specific subsets of the local epiphytic species pool according to the specific properties of the trees [43][44][45]. The four selected trees were chosen not only because of their different bark common enough in the laurel forest to justify their presence in each selected plot.
Dead and leaning trees were avoided, and only those that had a diameter at breast height (DBH) > 22 cm were sampled. To study lichen diversity, we used a frame of 50 × 10 cm subdivided into 5 quadrats of 10 × 10 cm. The uppermost edge of the frame was located at 1.5 m above ground level, adjusted up to a maximum of 2 m if the trunk was unsuitable at the desired height, following Refs [46,47] with some modifications in order to adjust these guidelines to the island territory (see Ref [48]). Thus, the number of quadrats studied by tree was 10 (5 in the north orientation and 5 in the south), and 80 quadrats were studied in each sampling plot. The survey was performed between September 2013 and March 2018.

Specimens Identification
At each quadrat of 10 × 10 cm, all lichens (macro-and microlichens) were collected and subsequently identified in the laboratory. All details of the identification methods were described in Ref [48]. Voucher specimens were deposited in TFC-Lich Herbarium of the University of La Laguna. The nomenclature of the lichen species mainly follows Refs [49,50]. Species names and authors of the taxa are outlined in Supplementary Table S1, they were omitted from the text. Dead and leaning trees were avoided, and only those that had a diameter at breast height (DBH) > 22 cm were sampled. To study lichen diversity, we used a frame of 50 × 10 cm subdivided into 5 quadrats of 10 × 10 cm. The uppermost edge of the frame was located at 1.5 m above ground level, adjusted up to a maximum of 2 m if the trunk was unsuitable at the desired height, following Refs. [46,47] with some modifications in order to adjust these guidelines to the island territory (see Ref. [48]). Thus, the number of quadrats studied by tree was 10 (5 in the north orientation and 5 in the south), and 80 quadrats were studied in each sampling plot. The survey was performed between September 2013 and March 2018.

Specimens Identification
At each quadrat of 10 × 10 cm, all lichens (macro-and microlichens) were collected and subsequently identified in the laboratory. All details of the identification methods were described in Ref. [48]. Voucher specimens were deposited in TFC-Lich Herbarium of the University of La Laguna. The nomenclature of the lichen species mainly follows Refs. [49,50]. Species names and authors of the taxa are outlined in Supplementary Table S1, they were omitted from the text.

Morphological Traits and Functional Groups
The species were classified according to the traits: (1) photobiont, (2) growth form, and (3) reproductive strategy. Additionally, the species were grouped in functional groups (4) according to their ecological requirements (tolerance to eutrophication, water requirements, solar irradiation, poleotolerance, and pH of the substrate) according to Refs. [50,51] ( Tables 2 and 3). Functional groups were considered as groups of species with a similar response to an environmental factor [51]. Following Refs. [51,52], we used the maximum value available for each ecological indicator. If no ecological data were available in the literature for a given species, the values for these ecological indicators were assigned based on expert assessments by Canarian lichenologists and our own field observations.

Data Analysis
The statistical procedures applied are described in depth in Ref. [48]. We used a Venn diagram to show the exclusive vs. shared species among the phorophytes Venny 2.1.0 [53]. In order to determine lichen diversity, we calculated the species richness (S) and lichen diversity value (LDV), following Refs. [46,47], and studied the significant value of the medians through the medians of the Kruskal-Wallis nonparametric test (H; p ≤ 0.05) and the Mann-Whitney test adjusted by the Bonferroni procedure (U; p ≤ 0.0167). The dependence relationship between the traits and lichen functional groups and laurel forest selected trees was studied with the chi-square test. To test if the ecological requirements of lichens (pH, solar irradiation, aridity, eutrophication, and poleotolerance) are similar among the trees, we used the community-weighted mean trait value CWM [54]. We built an abundance matrix with the plot, tree, and exposure, following Ref. [55], and computed a non-metric multidimensional scaling (NMDS) with 200 permutations and the Jaccard distance [56] to reveal the spatial order of lichen composition of the four compared trees, in two dimensions. To study the adequacy of the sample configuration in the NMDS, we analyzed the stress value [57]. Finally, to determine whether there were significant differences between the phorophytes, we ran an ANOSIM test [58], and to identify the lichen species that contributed most to the similarity (and dissimilarity) among the phorophytes, we used the SIMPER analysis [58].
Regarding the ecological conditions preferred by the recorded lichens, Figure 5 summarizes the responses of the exclusive species (or unique species, i.e., those lichens that only grow on a certain phorophyte) occurring on the studied trees concerning the pH, solar irradiance, aridity, eutrophication, and poleotolerance. The epiphytic lichens of Morella faya and Erica canariensis share pH substrate categories 2 ("on acid substrate") and 3 ("on subacid to subneutral substrate"), whereas the epiphytic lichens of Laurus novocanariensis and Ilex canariensis are mainly associated with substrate pH level 3 (Figure 5a). The response of epiphytic lichens to solar irradiation differs among all the studied phorophytes; the lichen composition of Morella faya is typical of "shaded situations" (category 2) and "sites with plenty of diffuse light but scarce direct solar irradiation" (category 3), followed by "sun-exposed sites, but avoiding extreme solar irradiation" (category 4), but some species present in Morella faya are typical of "sites with very high direct solar irradiation" (category 5). A similar response is observed in the epiphytic lichens of Ilex canariensis, with most lichens in categories 2 and 3, followed by category 4. The epiphytic composition of Laurus novocanariensis is mainly grouped in category 3, followed by category 4. Finally, Erica canariensis composition differs from that of other phorophytes, with lichen species typical of illuminated sites being more abundant (category 5, followed by categories 4 and, to a lesser extent, 3) (Figure 5b). We did not observe any species typical of a "very xerophytic" environment (category 5). The lichen composition of Morella faya, Laurus novocanariensis, and Ilex canariensis is shared between categories 2 ("intermediate between hygrophytic and mesophytic stands") and 3 ("mesophytic"), and the composition of Erica canariensis is shared among categories 3 and 4 ("xerophytic, but absent from extremely arid stands"), followed, to a lesser extent, by category 2 (Figure 5c). We did not find any species typical of "very high eutrophication" (category 5). The lichen composition of Laurus novocanariensis and Ilex canariensis has a similar response to eutrophication-typical of "very weak eutrophication" stand (category 2), followed by "no eutrophication" site (category 1) and "weak eutrophication" environment (category 3). The lichen composition of Morella faya is divided between category 1 (50% of the composition) and categories 2 and 3 (the other 50% of the composition). Finally, the lichen composition of Erica canariensis has a bimodal structure, with lichens typical of no eutrophication places (category 1) and species typical of "rather high eutrophication" sites (category 4) (Figure 5d). With regard to poleotolerance, Laurus novocanariensis composition is equally distributed among categories 0 ("species, which exclusively occur on old trees in ancient, undisturbed forests"), 1 ("occurring in natural or semi-natural habitats"), and 2 ("in moderately disturbed areas"). The lichen composition of Ilex canariensis is typical of poleotolerance categories 1 and 2, followed by category 0. The epiphytic lichens of Morella faya are linked to category 1, and, to a lesser extent, to category 0. Finally, the lichen composition of Erica canariensis has a bimodal structure, with maxima in categories 1 and 3 ("species occurring in heavily disturbed areas") ( Figure 5e).  Regarding the ecological conditions preferred by the recorded lichens, Figure 5 summarizes the responses of the exclusive species (or unique species, i.e., those lichens that only grow on a certain phorophyte) occurring on the studied trees concerning the pH, solar irradiance, aridity, eutrophication, and poleotolerance. The epiphytic lichens of Morella faya and Erica canariensis share pH substrate categories 2 ("on acid substrate") and 3 ("on subacid to subneutral substrate"), whereas the epiphytic lichens of Laurus novocanariensis and Ilex canariensis are mainly associated with substrate pH level 3 (Figure 5a). The response of epiphytic lichens to solar irradiation differs among all the studied phorophytes; the lichen composition of Morella faya is typical of "shaded situations" (category 2) and "sites with plenty of diffuse light but scarce direct solar irradiation" (category 3), followed by "sun-exposed sites, but avoiding extreme solar irradiation" (category 4), but some species present in Morella faya are typical of "sites with very high direct solar irradiation" (category 5). A similar response is observed in the epiphytic lichens of Ilex canariensis, with most lichens in categories 2 and 3, followed by category 4. The epiphytic composition of Laurus novocanariensis is mainly grouped in category 3, followed by category 4. Finally, Erica canariensis composition differs from that of other phorophytes, with lichen species typical of illuminated sites being more abundant (category 5, followed by categories 4 and, to a lesser extent, 3) (Figure 5b). We did not observe any species typical of a "very xerophytic" environment (category 5). The lichen composition of Morella faya, Laurus novocanariensis, and Ilex canariensis is shared between categories 2 ("intermediate between hygrophytic and mesophytic stands") and 3 ("mesophytic"), and the composition of Erica canariensis is shared among categories 3 and 4 ("xerophytic, but absent from extremely arid stands"), followed, to
The percentages of dissimilarity obtained through the SIMPER analysis were high in all cases, reaching 85%, although the highest percentage values were obtained among the comparisons of differentiated groups (92-98%) and the lowest among Erica canariensis vs. Morella faya (87%) and Ilex canariensis vs. Laurus novocanariensis (88%) ( Table 4).
The SIMPER analysis revealed discriminant species. Not all species contributed equally to the differences between the phorophytes. A list of characteristic species, with percentages of contribution in brackets, of each phorophyte studied is shown in Table 5.
the lichen composition of Erica canariensis has a bimodal structure, with maxima in categories 1 and 3 ("species occurring in heavily disturbed areas") ( Figure 5e).
The percentages of dissimilarity obtained through the SIMPER analysis were high in all cases, reaching 85%, although the highest percentage values were obtained among the comparisons of differentiated groups (92-98%) and the lowest among Erica canariensis vs. Morella faya (87%) and Ilex canariensis vs. Laurus novocanariensis (88%) ( Table 4). The SIMPER analysis revealed discriminant species. Not all species contributed equally to the differences between the phorophytes. A list of characteristic species, with percentages of contribution in brackets, of each phorophyte studied is shown in Table 5.  Some discriminant species (discriminating species of phorophytes according to the SIMPER analysis; Table 5) are common among phorophytes, this number being higher between Laurus novocanariensis and Ilex canariensis (four common discriminant species: Bacidia absistens, Lobaria macaronesica, Ricasolia virens, and Phlyctis agelaea), between Erica canariensis and Morella faya (two species: Chrysothrix candelaris and Platismatia glauca), and between Ilex canariensis and Morella faya (only Thelotrema lepadinum). On the other hand, two discriminant species are common to Laurus novocanariensis, Erica canariensis, and Morella faya: Parmotrema perlatum and Leucodermia leucomela. No discriminant species is common to all the studied phorophytes.
NMDS ordination resulted in a two-dimensional pattern with a stress value of 0.28 and revealed a trend in lichen species composition among the studied phorophytes. Laurus novocanariensis and Ilex canariensis appear on the right side of the plot, whereas Morella faya and Erica canariensis appear on the left side of the plot, primarily. However, three samples of Morella faya, corresponding to plots 7, 11, and 12, and one sample of Erica canariensis (plot 11), all from La Palma Island, are on the right side of the NMDS plot. Finally, two samples of Laurus novocanariensis, one of them from Tenerife (plot 6) and another from La Gomera (plot 13), are on the left of the NMDS plot ( Figure 6).
The species richness found in the Canarian Monteverde (165 species in total) varies slightly among the phorophytes, with values of 18.3/m 2 in Erica and Morella, 20.2/m 2 in Laurus, and 24.7/m 2 in Ilex. Some authors have described that trees with different bark characteristics have different epiphyte species. For this reason, a forest environment in which phorophytes with different barks coexist will accumulate a higher diversity of epiphytes than monospecific forest formations [48,70,71].
The species richness found in the Canarian Monteverde (165 species in total) varies slightly among the phorophytes, with values of 18.3/m 2 in Erica and Morella, 20.2/m 2 in Laurus, and 24.7/m 2 in Ilex. Some authors have described that trees with different bark characteristics have different epiphyte species. For this reason, a forest environment in which phorophytes with different barks coexist will accumulate a higher diversity of epiphytes than monospecific forest formations [48,70,71].
Although there are no previous studies in the Canarian Monteverde that allow us to compare our data with those studies, we were able to do so with the richness of epiphytic macrolichens identified in a study of the Madeiran monteverde, with a value of 1.4 epiphytic macrolichens/m 2 [72]. This value is more than 4 times lower than ours if we consider only the macrolichens (5.8 macrolichens/m 2 , 83 species of the 165 observed). It is very relevant to highlight the importance of the contribution of microlichens to the total specific richness. Studying bryophytes on different phorophytes (Laurus novocanariensis, Morella faya, and Erica canariensis) in Madeira [35], it was found that Laurus was the species tree that hosted the most epiphytic species, which agrees with our work in this sense. On the other hand, Ref [73] estimates a value of 5.3 epiphytic bryophytes/m 2 for Laurus in the Monteverde of Tenerife. This value is much lower than the lichen diversity that we observed on this same phorophyte. This same pattern was observed by Ref [74] in the monteverde of the Garajonay National Park (La Gomera island), where lichen diversity (318 species, 52% of them epiphytes) exceeded bryological diversity (199 species, 36.2% of them epiphytes). Although exceptions exist [75], similar behavior has been observed in other forest environments worldwide [76][77][78].
Considering the high frequency of taxa with common traits (biotype, photobiont, etc.), as well as the diversity of the genera and families present in the four phorophytes studied, there seems to be a disjunction in the Canarian Monteverde between the lichens present in Ilex and Laurus and those that grow in Morella and Erica, as revealed by the NMDS analysis ( Figure 6). Our results are in accordance with the patterns found for bryophytes in previous studies in the Monteverde of Macaronesia, both on the island of Tenerife [32,33] and Madeira [35].

Functional Traits
Functional traits have been used as bioindicators of forest attributes of interest for forest characterization and management [79][80][81]. However, some authors point out that categorizing lichens according to their morphological traits rather than the species, genera, or families to which they belong could lead to erroneous conclusions [82]. For this reason, we will take into account both the functional traits of lichens and their identity.
The study areas were established in homogeneous (meso)climatic zones. Thus, it is expected that the differences found between the two groups of phorophytes described (Morella + Erica and Laurus + Ilex) are due to intrinsic characteristics, such as: tree diameter and age, presence of latex and other epiphytes, chemical composition, pH, stability, roughness, thickness and water-holding capacity of the bark, number and size of lenticels, and cortical runoff [27,[83][84][85][86][87][88][89][90][91][92][93][94]. Thus, we can define Morella and Erica as two trees with acid pH, low stemflow funneling ratio values, small leaf area [95,96], and with a thick, rough bark; and Ilex and Laurus as two trees with neutral pH, high stemflow funneling ratio values, larger leaves [95,96], and smooth and thin bark. None of these trees have a flaking bark or latex.
In terms of the most frequent and diverse taxa found in the studied phorophytes, we observed that crustaceous lichens are very common in Ilex. Erica is the only phorophyte in which the foliaceous biotype predominates over any other. Crustaceous species richness has been negatively related to tree bark roughness by some authors [30,31]. However, other studies have shown that different crustaceous lichens show different responses to this phorophyte feature [71]. As observed by Refs [30,31], roughness could explain the greater diversity found in Ilex (especially) and Laurus, since both trees have a smooth bark compared to Morella and Erica, which have a rough bark. On the other hand, foliaceous species with large thalli, mainly from the Parmeliaceae family, show a high biodiversity in Erica and Morella, where they are more frequent, as well as in Laurus. The dichotomy between Erica + Morella and Laurus + Ilex, previously mentioned, is further emphasized when studying the dimorphic biotype, represented only by the genus Cladonia. It is found exclusively on Erica and Morella. Cladonia has generally been associated with acid substrates, where it predominates [97,98]. This could explain the presence of Cladonia exclusively in Erica (pH = 5.2 ± 0.7) and Morella (pH = 4.5 ± 1.2) and its absence in Ilex (pH = 6.4 ± 0.4) and Laurus (pH = 6.3 ± 0.6) [95], although variations of this pattern may exist [99]. The Cladonia species identified are mostly found in heathlands and acid substrates, both in the monteverde and in other forest environments [69,[100][101][102]. In Erica and Morella, there is also a high frequency of Chrysothrix candelaris, a leprose lichen that has been related to acid substrates, rough bark, and environments with high luminosity, both in the Monteverde and in other forest environments [27,69]. However, it has been described as a photophobic species, frequent in acidic and rough crusts [103], which does not tolerate eutrophication [50] and has been observed in well-preserved environments of the Canarian Monteverde [101].
When studying the most frequent species, as well as the most diverse genera and families, we only found cyanolichens (Sticta canariensis cyanomorphotype, Crocodia aurata, Sticta, and Leptogium) and tripartite lichens (with green algae as the main photobiont and cyanobacteria as the secondary photobiont: Ricasolia virens and Lobariaceae) in Ilex and Laurus. However, no cyanolichens were found among the most frequent species in Erica or Morella. No representative of this group was found in Morella. Although there seem to be exceptions [104], it is generally accepted that cyanolichens need liquid water to photosynthesize [105]. Stemflow funneling is defined by incident rainfall, cortical runoff volume, and diameter at breast height (DBH), but it is also influenced by the roughness of the trunk and branches, as well as by the angle they form with respect to the horizontal. The stemflow funneling ratio data obtained by Ref. [95] for the different trees (with values of 42.5 for Ilex, 28.5 for Laurus, 22.8 for Erica, and 12.7 for Morella) support our results. The higher richness of lichens with Trentepohliaceae on Ilex compared to the rest of the studied phorophytes could be explained by the high number of crustaceous species that develop on this tree. All lichens with this photobiont have this biotype. Finally, the predominance of lichens with chlorococoid green algae found in the monteverde is simply a consequence of their higher abundance in nature [17].
Regarding reproductive strategies, sexual reproduction predominates in the monteverde, followed by multiplication by soredia, isidia, a conjunction of both (soredia and isidia), and fragmentation of the thallus. This pattern has also been described in other forest environments [48,106,107]. However, a more detailed analysis reveals, again, a different pattern among the studied phorophytes. The occurrence of lichens with sexual reproduction is particularly significant in Ilex and Laurus, while in Morella and Erica, sexually reproducing species and those that multiply asexually (usually by soredia) co-dominate. Considering that the trees in a given study plot are within a few meters of each other, the lower abundance of sorediate lichens on Ilex and Laurus compared to Morella and Erica cannot be explained solely based on the higher difficulty of dispersal of these propagules compared to spores. It has been shown that soredia can reach more than 200 m [108]. Therefore, there must be some feature of the phorophytes that explains this finding. A hypothesis could be that it is related to the bark roughness. Ref. [109] concluded that bark microtopography significantly influenced the establishment and survival of lichens dispersed by soralia; survival was poor on smooth bark compared to survival on rough bark. The smaller size of the spore compared to soredia could explain why species with soralia are less frequent on trees with a smooth bark, where the attachment is more complex compared to that on a rough bark. The same reasoning has been used in the case of bryophyte and fern spores to explain the higher probability of adhesion to a rough surface vs. a smooth surface [110][111][112].
The presence and abundance of some epiphytic lichens in forests are related both to the characteristics of the forest itself and to the ecological requirements of the lichen species [83]. Thus, if we know the ecological requirements of the species inhabiting a forest, we can obtain information about the forest indirectly.
Epiphytic lichens are influenced by macro-, meso-, and microclimatic factors, as well as by the physical and chemical features of the substrate where they grow. Because of these requirements, several authors have used lichens as bioindicators of both polluted and well-preserved environments [113,114]. Most authors adopt the proposal of Ref. [50], using their classification in various regions of Mediterranean influence [48,51,[115][116][117]. However, we are not aware of previous studies that have used this methodology to study the specificity of epiphytic lichens and their phorophytes.
Due to the architecture and structure of the studied tree species, the morphology and size of leaves, the characteristics and nature of the bark affecting the water-holding capacity, runoff, etc., a differential response of lichens to solar radiation, aridity, and even pH would be expected. Regarding pH, it should be noted that lichens found exclusively on Laurus and Ilex have an affinity for subacid to subneutral substrates, and these trees are the only ones with lichens species with an affinity for slightly basic substrates. However, the epiphytic lichens growing only on Morella and Erica are associated with acidic and subacid to subneutral substrates, with no exclusive lichen species having an affinity for (slightly) basic substrates. This is consistent with the previous comments on the pH characteristics of the four phorophytes and the occurrence of particular groups of lichens associated with this character.
In terms of the response of lichens to solar irradiation, we expected to obtain a certain homogeneity due to the similar characteristics of the canopy of the different localities and the fact that all are evergreen and broadleaf trees, with the exception of Erica, which has substantially different leaves. In the case of Ilex, Laurus, and Morella, we expected to obtain a significant proportion of poorly light-tolerant lichens, typical of shady environments, such as evergreen forests like the Canarian Monteverde (category 1 sensu Ref. [50]). We did not find a common or distinctive pattern in the trees studied, nor was our hypothesis regarding ombrophilous lichens fulfilled (Figure 5b). However, we were able to detect certain trends. Lichens growing on Laurus and Ilex seem to have a certain affinity for environments with a lot of diffuse light but little direct solar irradiation, while the response to light in the epiphytes of Morella and Erica is completely opposite and different from Laurus and Ilex. In Morella, sciaphilous species dominate, with almost 50% of its epiphytes in category 2, while Erica is dominated by species with very high direct solar irradiation requirements, with more than half of its species in category 5. Both Morella and Erica are heliophilous and primocolonizing species of the Canarian Monteverde [118]. However, Erica does not usually take part in mature Monteverde communities, while Morella remains, even being part of the mature (climax) phase of the forest [119].
On the other hand, the leaves of Erica are by far the smallest and narrowest of the four trees analyzed [95,96]. Therefore, it would be expected that higher solar radiation penetrates under its canopy, and consequently, more lichens with high light requirements would grow on its bark, as is the case here. The leaves of Morella, although some authors considered them small compared to those of Laurus or Ilex [95], are much wider and longer than those of Erica. This would explain the higher abundance of lichens related to shady environments with respect to Erica but less in relation to Laurus or Ilex.
As a response to the aridity, it could be hypothesized that the epiphytic lichens of all monteverde trees studied have ecological indicator values for this indicator, which are typical of hygrophytic environments (low values), as it is a forest linked to high humidity [120]. Erica epiphytes tend to tolerate a higher degree of aridity than expected (Figure 5c). This could be explained partly by the solar irradiance data mentioned above and also by the low channeling values of this tree [95]. In the case of Morella, although the channeling values are not high [95], the water retention capacity of the bark of this tree, particularly in the larger and corky-looking specimens, could be playing a significant role. This could explain why this phorophyte has the highest percentage of species classified as slightly hygrophytic. Overall, Laurus and Ilex could be classified as two trees with epiphytic lichens typical of a mesophytic to slightly hygrophytic environment. Considering the four phorophytes analyzed in this study, this forest could be defined as a mesophytic to slightly hygrophytic forest environment.
For the eutrophication and poleotolerance, low values were expected for the four phorophytes analyzed, as it is a primary forest. The values for Ilex and Laurus are in accordance with those expected, with a high proportion of lichens occurring exclusively on old trees in ancient, undisturbed forests or natural (or semi-natural) habitats. Consequently, they are mostly non-resistant to eutrophication or resistant to a very weak eutrophication (Figure 5d,e). Morella also has a unique diversity of epiphytic lichen diversity with low eutrophication and poleotolerance values. The case of Erica is particular in this respect, showing the highest values of eutrophication and poleotolerance. These high values are related to the presence of a single species: Amandinea punctata. It is responsible not only for the maximum observed in these indicators but also has a great impact on the solar irradiation values. This species has only been found in one plot on Erica on the island of La Palma but with an astonishing abundance. It was reported for the first time for the Canary Islands in the Caldera de Taburiente National Park (La Palma) on various substrates [121], but it has also been collected in the monteverde of the Garajonay National Park (La Gomera) on rocky outcrops in open and somewhat degraded areas of Morella and Erica forest [20], and in El Teide National Park in Tenerife [122]. The diversity of the environments in which this species has been reported shows its wide ecological value and it should not be considered in any case as a bioindicator species of the Canarian Monteverde. As previously stated, we always used the maximum ecological value, and therefore, the analysis may be affected in this case.

Lichen Composition
The low proportion of shared lichens among all the studied phorophytes compared to the diversity found exclusively in each phorophyte, suggests a possible lichen-phorophyte specificity. The specificity of epiphytic lichens and their phorophytes is a controversial issue, and it does not seem that the same patterns occur in all forest ecosystems. Thus, while some authors report some specificity [28,29], others do not see a clear pattern [30,123], and some authors even suggest that the results obtained in this regard in plurispecific forests should be taken with some cautiousness [42,124].
The high species richness shared between Erica and Morella on the one hand, and between Laurus and Ilex on the other, suggests a possible affinity of lichens for the trees that comprise these two subsets. In this case, this pattern seems to be consistent with what has been reported by other authors in tropical forests. Several authors have found similar patterns, i.e., affinity of lichen species for groups of phorophytes, and have explained them based on the physical and chemical features of the trees and their bark (pH, roughness, water retention capacity, tree perimeter, etc.) [30,31,85,125]. It should also be noted that the two aforementioned subgroups (Laurus novocanariensis + Ilex canariensis and Morella faya + Erica canariensis) are the same as those observed by Refs. [32,70] when studying epiphytic bryophytes in the monteverde of La Gomera.
According to Ref. [30], the specificity of epiphytes to phorophytes is not beneficial for lichens in forests with a plurispecific tree composition, as it decreases the probability of finding and establishing lichen diaspores with a suitable phorophyte. This can be an issue when the diaspores are asexual propagules, as they are larger and heavier than sexual spores and have lower dispersal facility [126][127][128]. This has implications for the conservation of monteverde, particularly on islands such as Tenerife, where the forest is fragmented [2], and should be considered in ecological restoration actions in the Canarian Monteverde.

Conclusions
No specificity of Canarian Monteverde lichens concerning the tree species analyzed was observed. However, an affinity toward two well-defined groups of phorophytes was identified: Erica + Morella and Laurus + Ilex. This differentiation is clear, considering the specific diversity and the number of lichen species shared by these two groups. These data are also supported by the patterns of functional traits of lichens that comprise these two groups. The differences found may respond to the physico-chemical features of the tree (leaf type and canopy structure, roughness, pH, and thickness of the bark, among others). Further studies would be necessary to elucidate the importance of each of these factors separately. The results presented will be useful to improve the ecological restoration efforts in the Canarian Monteverde.