Unravelling the diversity of the lichen genus Porina (Porinaceae) in Mauritius

Background and aims – Despite the publication of a recent checklist for Mauritius, a small archipelago in the south-western Indian Ocean, our knowledge of the lichen flora of this region remains incomplete. The present contribution is part of an ongoing study of lichen diversity of the islands of Mauritius and Rodrigues. It aims to unravel the diversity of the genus Porina s.l. on these islands and to improve the taxonomy and the phylogeny of the family Porinaceae following the morphological study and the sequencing of recently collected material. Material and methods – Lichens were collected in 2016 and 2019 in a range of environments, from lowland dry habitats up to dense evergreen upland forests on the islands of Mauritius and Rodrigues. In total, 85 samples of Porina were examined using light microscopy. A molecular study using mtSSU sequences was performed in order to investigate the phylogenetic position of the species of Porina occurring in the study area. Key results – The analyses revealed the presence of 23 species of Porina s.l. in Mauritius, while only four were previously accepted for this country. Five new species are described: Porina covidii, P. griffithsii, P. mauritiana, P. nuculoides, and P. rupicola, while two sterile species remain unidentified. The mtSSU phylogenetic tree includes 19 of the 23 species occurring in Mauritius. These species are dispersed throughout the phylogenetic tree, in at least eight main and strongly supported lineages, suggesting multiple colonizations of Mauritius. A new lineage only including Paleotropical and mainly foliicolous species (P. epiphylloides, P. longispora, P. mauritiana, and P. mazosioides) is revealed. Pseudosagedia crocynioides, a species that was previously known only from Florida, is newly reported from Mauritius, highlighting a remarkable disjunct distribution. Photos of each taxon are provided, along with an identification key to all species. Conclusion – The study of the genus Porina in Mauritius highlights a rich flora. The true diversity of Porina species in the south-western Indian Ocean is certainly underestimated and in need of further investigation, while a taxonomic-phylogenetic revision of tropical Porina at a broader geographical scale is highly desirable. The identity of several Mauritian specimens remains uncertain despite the use of molecular data.


INTRODUCTION
Porina Ach. is a diverse and widespread genus with ca 75% of the species occurring in humid-subtropical and tropical latitudes, the remainder inhabiting mainly temperate regions. The genus forms one of the main elements of the epiphytic lichen communities in humid tropical forests. The taxa are usually substrate-specific, with nearly half of the species colonizing bark, one-third leaves, ca 20% rocks, and a few bryophytes or soil (McCarthy 1993(McCarthy , 2003. The genus is mainly recognized by the combination of the following characters: thallus crustose with a trentepohlioid photobiont, usually lacking secondary metabolites; ascomata perithecioid; paraphyses unbranched to sparingly branched; hymenium IKI-; asci unitunicate, 8-spored, IKI-; ascospores colourless, thin-walled, narrowly ellipsoid, fusiform to filiform, usually transversely septate, rarely muriform. The family Porinaceae includes 359 accepted species . Molecular phylogenies using mainly mitochondrial SSU rDNA (mtSSU) sequences recovered the genus Porina as paraphyletic, with the genera Myeloconis P.M.McCarthy & Elix and Trichothelium Müll.Arg. nested within Porina (Grube et al. 2004;Baloch & Grube 2006;Nelsen et al. 2014). Yet, these genera were maintained because they are well characterized: Trichothelium by perithecia with setae (Lücking 2008) and Myeloconis by yellow to orange medullary phenalenones (McCarthy & Elix 1996;Ernst-Russell et al. 2000). Attempts to define further smaller natural entities mainly used the thalline and perithecial morphology (e.g. presence of oxalate crystals in the thallus, pigmentation of the perithecial wall, degree of thallus cover of the perithecia, presence and type of setae on the perithecia, ascus and ascospores types), as well as environmental preferences and distributional trends (e.g. Hafellner & Kalb 1995;Harris 1995;Malcolm & Vězda 1995). Two new genera were even recently described, based mainly on the thallus type in combination with molecular data: flabelliform squamules for the genus Flabelloporina Sobreira, M.Cáceres & Lücking and placodioid thallus with a fenestrate anatomy for the genus Saxiloba Lücking, Moncada & Viñas (Sobreira et al. 2018;Lücking et al. 2020). However, a splitting of the Porinaceae in small generic entities has not been widely accepted because characters are often shared between genera and are highly variable (e.g. McCarthy & Malcolm 1997;Lücking 2008) so that a broader paraphyletic concept of Porina is currently in use. A generic rearrangement or at least a subgeneric arrangement of the Porinaceae could be possible but requires analyses supported by molecular data and a more complex combination of morphological characters instead of the previous rather schematic treatments (Lücking 2008: 271). A possible generic concept for Porinaceae was outlined using a molecular phylogeny, but the results were based on a single marker, the mtSSU (Sobreira et al. 2018;Lücking et al. 2020).
A first checklist of lichens and lichenicolous fungi in the Republic of Mauritius was recently published and includes 216 accepted species (Diederich & Ertz 2020). The genus Porina is represented only by four species, viz. P. epiphylla (Fée) Fée, P. florensii Diederich & Ertz (newly described within the checklist), P. nucula Ach., and P. tetracerae (Ach.) Müll.Arg. var. tetracerae. Four additional taxa (P. americana Fée, P. chlorotica (Ach.) Müll.Arg., P. innata (Nyl.) Müll.Arg., P. mastoidea (Ach.) Müll.Arg. agg.) have been reported from Mauritius, but these were not accepted in the checklist, either because no relevant herbarium material has recently been examined, or because previous records are dubious or erroneous. With the aim of completing our knowledge of the lichens and lichenicolous fungi of the islands, about 2000 specimens were collected in Mauritius and Rodrigues during a field trip in August and September 2019. These include 79 specimens of Porina, a genus that obviously represents a species-rich group in the study area. The genus Porina has long been neglected, but important taxonomic works published in recent decades facilitate the identification of tropical specimens. A world-wide catalogue of the Porinaceae is available (McCarthy 2003(McCarthy , 2013) and important treatments have been published for the saxicolous species in the Southern Hemisphere (McCarthy 1993), the corticolous species in Australia (McCarthy 1994a(McCarthy , 1994b, the Porina epiphylla group (Lücking & Vězda 1998), and the foliicolous species in the Neotropics (Lücking 2008). Several regional identification keys are also available for northeastern Brazil (Cáceres 2007), Florida (Harris 1995), the Indian Subcontinent (Awasthi 1991), Japan (Harada 2015(Harada , 2016, and Macaronesia (Sérusiaux et al. 2007).
This paper aims to (1) unravel the diversity of the genus Porina in Mauritius by the identification of recently collected specimens, (2) update the phylogeny of the genus by sequencing all Mauritian species, and (3) describe new taxa resulting from our morphological and molecular investigations.

Morphology and chemistry
Voucher specimens are deposited in the herbaria BR, MAU, and TAN. The external morphology was studied and measured using an Olympus SZX12 stereomicroscope. Macroscopic images were captured with a Keyence VHX-5000 digital microscope and a VH-Z20R/W/T lens. Handcut sections and squash preparations of ascomata and thalli were mounted in water, a 5% aqueous potassium hydroxide solution (K), or in Lugol's iodine solution (1% I2) without (I) or with K pretreatment (KI) and studied using an Olympus BX51 compound microscope. The K and I solutions were by default added to a water mount by running a drop under the edge of the coverslip. Calcium oxalate was identified by adding 25% H 2 SO 4 and the refractive ring structure of asci by using an ammoniacal Congo Red solution (1%). Measurements from ascomatal sections (e.g. hymenium, asci, ascospores) refer to dimensions in water. Measurements of asci are reported as minimum-maximum followed by the number of measurements (n), as only few were measured. Measurements of ascospores do not include the perispore (= gelatinous sheath) and are reported as (minimum-) SD -SD (-maximum), followed by the number of measurements (n), or as minimum-maximum when few ascospores were measured, and the values usually rounded to the nearest 0.5 µm. Micrographs were prepared using an Olympus BX51 compound microscope fitted with an Olympus SC50 digital camera. Colour reactions of the thallus were studied using K, household bleach (C), K followed by C (KC), crystals of paraphenylenediamine dissolved in ethanol (PD) and long-wave UV (366 nm). Lichen secondary metabolites were investigated using thin-layer chromatography (TLC) in solvent A (Orange et al. 2010), but only a few specimens were tested because Porina usually lacks acetone-soluble secondary metabolites detectable by TLC.

Molecular techniques
Well-preserved specimens lacking any visible symptoms of fungal infection, either freshly collected (less than one month, except for Porina borreri Ertz 12130, P. cf. malmei Ertz 25249, and P. perminuta Diederich 18732 that were respectively a 12-year-old, a 14-month-old, and a one-yearold herbarium specimen) or kept in the freezer and frozen less than one month after collection, were used for DNA isolation. Hand-cut sections of the perithecia and the thallus, or a small number of soredia (Porina sp. B Ertz 23618) or isidia (Porina florensii Ertz 24283, P. griffithsii Ertz 23632, P. sp. A Ertz 23584) were used for direct PCR as described in Ertz et al. (2015). The material was placed directly in microtubes with 20 μl H 2 O. Amplification reactions were prepared for a 50 μl final volume, as detailed in Ertz et al. (2018). For all species, we tried to amplify a targeted fragment of ca 0.8 kb of the mtSSU rDNA using primers mrSSU1 and mrSSU3R (Zoller et al. 1999). For Pseudosagedia crocynioides, a fragment of about 0.6 kb of the nITS rDNA (ITS1 + 5.8S + ITS4) was amplified using primers ITS1F and ITS4 (White et al. 1990). Both strands were sequenced by Macrogen ® using the amplification primers. Sequence fragments were assembled with Sequencher v.5.4.6 (Gene Codes Corporation, Ann Arbor, Michigan). Sequences were subjected to 'megablast' searches to verify their closest relatives and to detect potential contaminations.

Phylogenetic analysis
Partial mtSSU sequences were newly generated for 36 specimens of Porinaceae, belonging to 21 species. All were from Mauritius, except two that were sequenced for comparison with a newly described species (see under P. covidii in the taxonomy section below): Porina borreri (Trevis.) D.Hawksw. & P.James from Belgium and P. cf. malmei P.M.McCarthy from Madagascar (supplementary file 1). Eighty-nine additional sequences were retrieved from GenBank ( fig. 1). The mtSSU matrix of Porinaceae consisted of 125 terminals and 594 unambiguously aligned sites. The RAxML tree obtained is shown in fig. 1. The main wellsupported lineages are in accordance with the results obtained by Sobreira et al. (2018) and Orange et al. (2020). The nodes of the backbone of the tree are mainly poorly supported, not providing informative phylogenetic structure for the family. The genus Porina is recovered as paraphyletic, lineages being intermingled with those of the genera Flabelloporina, Myeloconis, Saxiloba, and Trichothelium. The newly prepared sequences are recovered in no less than eight main, strongly supported lineages. Notes -New to Mauritius. Pantropical. Our material fits well with the concept of P. africana by perithecial verrucae ca 0.5-1 mm diam., a thallus with a black basal layer and 7-8(-10)-septate ascospores (33-)40-54.5(-63) × 5-6.5(-7) µm (n = 59). It deviates from the description of McCarthy (2001, "perispore not apparent") by ascospores often having a thin gelatinous sheath ca 1 µm wide. The periostiolar area is often dark brown to black as in the holotype of P. africana, but perithecia in Ertz 24306 lack a dark pigmentation suggesting that this character state may be variable within the species. The mtSSU sequences of Ertz 23957 and Ertz 24298 are identical and form a sister clade to two sequences obtained from Neotropical specimens of Porina tetracerae (KJ449309 from Costa Rica, KJ449314 from Panama) but ◄▲Figure 1-A-B. Phylogeny of the Porinaceae based on a data set of mtSSU sequences that resulted from a RAxML analysis. Maximum Likelihood bootstrap values are shown above or near the internal branches. Internal branches considered strongly supported are represented by thicker lines. The newly sequenced samples are in bold and their names followed by collecting numbers of authors, which act as specimen and sequence identifiers. Lineages corresponding to sequences obtained from Mauritian specimens are highlighted. The length of the branch represented by dashed lines was reduced by 50% for layout reasons. appear distinct from them because of the long branches. Porina tetracerae differs by somewhat smaller perithecial verrucae (0.36-0.8 mm diam.) and shorter ascospores (24-44 µm) (McCarthy 2001). When the black basal layer of the thallus is apparent, specimens recall P. internigrans, but the ascospores of the latter are longer and much broader. Unfortunately, no sequences of P. africana (originally described from Angola) are available on GenBank for comparison with our data. James by its black perithecia not immersed in distinct thallus verrucae and its mainly 7-septate ascospores but differs by having asci without an apical ring like structure and (5-)5.5-7.5(-8) µm broad ascospores that are at most 5.5 µm broad in P. borreri. Description -Thallus mostly epiphloeodal, continuous to rimose, smooth to slightly rugulose, matt, pale to dark greyish green to brownish green, thin ca 25-90(-125) µm thick, without crystals of calcium oxalate, without a black basal layer; prothallus non-apparent; isidia and soralia absent. Photobiont trentepohlioid; cells spherical ca 5-8 µm diam. or slightly elongated 8-10 × 5-7 µm. Ascomata perithecioid, solitary, rarely two or three contiguous, dispersed, subglobose, black, smooth, 150-260(-280) µm diam., the base to ca 2/3 of the perithecia immersed in the thallus, without forming distinct thallus verrucae; ostiole apical, inconspicuous; crystallostratum absent. Proper excipulum hyaline to pale brown, fused with the involucrellum in ± the upper half to 2/3, ca 15-20 µm below. Involucrellum dark reddish brown to carbonized, ca 30-40 µm thick, often extending down to the base of the perithecium, except in perithecia that are strongly immersed in the thallus where the involucrellum might extend only to 1/3 of the perithecia, Kor becoming darker (reddish brown pigment disappearing). Hamathecium hyaline, not inspersed, of thin, simple, 1-1.5 μm diam. paraphyses, 110-160 µm tall; subhymenium pale fawn, 7-13 µm thick. Asci cylindrical-clavate to ± fusiform, I-, ca 80-105 × 15-17 µm (n = 4), 8-spored; ascus apex rounded, without a ring structure. Ascospores hyaline, I-, transversely (5-)7-septate, elongate-fusiform to ± oblong, (20-)22.5-28.5(-32) × (5-)5.5-7.5(-8) µm, ratio L/B (3.5-)3.5-4(-4.5) (n = 25); gelatinous sheath usually poorly visible, ca 2-3 µm thick. Pycnidia not observed. Chemistry: thallus K-, C-, KC-, PD-, UV-. TLC not performed. Distribution and ecology -The species is known from two localities in Mauritius, where it inhabits the bark of trees in rather open forests including parklands, at low elevation (ca 280-360 m). It is probably also present in the Seychelles (Schumm & Aptroot 2010, sub. P. borreri (Trevis.) D.Hawksw. & P.James). Etymology -The epithet refers to the Coronavirus disease of 2019 . The new species was described during the COVID-19 pandemic. Notes -Our material is similar to specimens identified as Porina borreri from the Seychelles (Schumm & Aptroot 2010). However, P. borreri differs from our Mauritian specimens in having asci with an apical ring structure and distinctly narrower ascospores: (3-)4.5-5(-5.5) µm for Great Britain and Ireland (Orange et al. 2009) and 3-5 µm for continental Europe (Sérusiaux et al. 2007). Moreover, P. borreri and the new species are not related in our phylogenetic analyses ( fig. 1), clearly supporting two different taxa. Among the saxicolous species keyed out by McCarthy (1993) for the Southern Hemisphere, Porina malmei P.M.McCarthy, described from Brazil, is similar in size to the perithecia and ascospores, but it differs in its saxicolous habit, a medium brownish grey thallus and a brown-black proper excipulum. We recently collected and sequenced a saxicolous specimen similar to Porina malmei in Madagascar, although our specimen differs in having a greenish thallus. In our phylogenetic tree, the latter is the sister species to P. covidii, but is phylogenetically distinct (long branches). Porina subchlorotica (Nyl.) Müll.Arg., known only from New Caledonia, is also similar but differs by larger (ca 300-400 µm diam.) perithecia and narrower ascospores (25-30 × 5-6 µm) (Nylander 1868 Harris (1995) for Florida and Cáceres (2007) for north-eastern Brazil, Porina cestrensis (Tuck.) Müll.Arg. (sub. Trichothelium cestrense) differs in having much longer ascospores, (32-)38-50 × 5.5-7.5 µm. Using the key of the corticolous and saxicolous species of Porina of Japan (Harada 2015(Harada , 2016, our material would be identified as P. guentheri (Flot.) Zahlbr., but that species differs by its saxicolous habit, larger perithecia (0.15-0.     (2001) already reported intermediate and often unidentifiable specimens, P. farinosa differing from P. eminentior only by its perithecia that tend to be more prominent and by its larger and more divided ascospores. In our material, the perithecia are weakly prominent, 0.65-0.8 mm diam., and better fit those of P. eminentior. The ascospores are ca 16-23 × 3-6 septate and 69-107 × 20-34 µm, and better fit those of P. farinosa by being slightly broader and having more longitudinal septa according to the descriptions in McCarthy (2001): 12-22 × (1-)2-3(-4) septate, 42-109 × 13-29 µm in P. eminentior; 17-30 × (1-)2-4(-6) septate, 58-148 × 18-41 µm in P. farinosa. Therefore, we attribute our material to P. eminentior with some hesitation. In our phylogenetic tree, specimen 23493 groups with P. exocha (Nyl.) P.M.McCarthy and P. farinosa, but appears different from both. Porina exocha differs by having larger perithecial verrucae (0.7-1.5 mm) and larger ascospores (68-168 × 20-45 µm) with a perispore with persistent hyaline apical caps (McCarthy 2001). Unfortunately, no sequences of P. eminentior were available on GenBank.  Santesson (1952: 234). Notes -Porina epiphylla is a common and widely distributed species in the tropics. It has been reported from Mauritius on the fern Acrostichum obductum (S F74625) by Santesson (1952: 238). Our specimen 24345 is minute with perithecia having a ± dark brown periostiolar area. It probably does not belong to P. epiphylla s.s., but possibly to P. atriceps (Vain.) Vain. The latter taxon was regarded as a variety of P. epiphylla by Santesson (1952), but accepted as a distinct species by Lücking & Vězda (1998). In our specimen, the photobiont cells are angular-rounded, irregularly arranged, and the ascospores are 7-septate, 28-35(-38) × 3-4 µm. However, our specimen is too scanty, and more material is needed to resolve its identity and possibly confirm the presence of P. atriceps in Mauritius. The second specimen (23526) is more typical for P. epiphylla but has ascospores of 31-36 × 3.5-4 µm, thus slightly longer than those mentioned by Lücking & Vězda (1998;26-32 µm long). It might better fit P. karnatakensis Makhija, Adawadkar & Patwardhan, but that species should have a black dot around the ostiole although becoming paler in old perithecia (Lücking 2008). In our phylogeny, the mtSSU sequences of our specimens do not group together, suggesting two different taxa, nor with those of the large clade of P. epiphylla from the Neotropics, indicating that our specimens belong to other taxa and that P. epiphylla s.s. might be absent from Mauritius. More material is needed to resolve the identity of the Mauritian taxa, as well as further molecular studies of the Porina epiphylla-group. Notes -New to Mauritius. Palaeotropical, known notably from tropical East Africa and the Comoros (Lücking & Vězda 1998). Our specimen is typical for the species in having a verrucose thallus, small perithecia (0.2-0.35 mm diam.) and small 7-septate ascospores, ca 25-28 × 3-3.5 µm. Notes -Endemic to Mauritius. This species was recently described as fertile and non-isidiate (Diederich & Ertz 2020). Fertile specimens with isidia were collected by us in Brise Fer Forest, and our phylogeny supports their inclusion in P. florensii. The mitochondrial sequence of the isidiate specimen (Ertz 24283) differs from the non-isidiate ones by only three nucleotides. Our morphological and phylogenetic results suggest that isidia do not represent an important taxonomic character at the species level in the genus Porina.
Pseudosagedia crocynioides was previously known only from Florida where it inhabits trunks of Taxodium in swampy forests. No significant morphological differences between our specimens and the original description of P. crocynioides were found. In the latter, ascospores were described as being slightly longer than in our specimens, "20-30 × 3-4 µm" in the Latin diagnosis but "20-35 × 2.5-3.5 µm" in the English description, and 3(-6)-septate. Asci were also slightly larger, 95-110 × 8-9 µm. As the original description is rather short and as the disjunct distribution is surprising, a detailed description of the Mauritian material is provided above. In the absence of molecular data from the populations of Florida, we cannot exclude the possibility of having two distinct species despite the similar morphology.

DISCUSSION
The genus Porina is diverse on the islands of the southwestern Indian Ocean, with 46 species currently known from the region (Seaward & Aptroot 2009;Schumm & Aptroot 2010;van den Boom et al. 2011;Aptroot 2016;Diederich et al. 2017; this study). Our revision of Mauritian material has highlighted a rich flora, with 23 species, five of which are described as new. The five Porina species found in the island of Rodrigues (P. covidii, P. cf. eminentior, P. griffithsii, P. nucula, and P. rupicola) are also present in Mauritius. This might be explained by the fact that these islands are close to each other (ca 574 km apart) and that Rodrigues is the smaller island (109 km 2 of land area vs 1865 km 2 for Mauritius) with the highest point (Mt Limon) reaching only 398 m in altitude (vs 828 m at Piton de la Petite Rivière Noire in Mauritius). Moreover, all pristine forests in Rodrigues have been destroyed by past human activity, unlike Mauritius that retains ca 2% of its native forest vegetation (Thébaud et al. 2009;Baider et al. 2010).
The genus Porina s.l. is well-represented on the other islands of the south-western Indian Ocean. Thus, 16 species have been reported from Réunion (van den Boom et al. 2011), seven of which are shared with Mauritius; 15 species are known from Madagascar (Aptroot 2016, and P. cf. malmei newly reported here), of which one-third are shared with Mauritius and the same proportion with Réunion; only one species (P. epiphylloides Vězda) was reported from the Comoro Islands (Lücking & Vězda 1998) but no checklist is available for this archipelago; 24 species are known from the Seychelles (Seaward & Aptroot 2009;Schumm & Aptroot 2010;Diederich et al. 2017), of which nine are shared with Mauritius. Only three species (P. epiphylla s.l., P. internigrans, and P. nitidula) are shared between Mauritius, Réunion, Madagascar, and the Seychelles. Two species (P. atrocoerulea Müll.Arg. and P. lucida R.Sant.) are present in the three latter, but are absent from Mauritius. These numbers do not include Trichothelium species, none of which have been collected in Mauritius.
The real species diversity of the genus in this region of the Indian Ocean is certainly underestimated. The lichen flora of the Comoros archipelago is largely unknown, and the checklist of Madagascar is only preliminary. In Mauritius, several sites known for their rich and original biodiversity were not visited by us for our revision of the genus, such as the Bambou Mountain Range, the Corps de Garde, and several offshore islets (e.g. Île aux Aigrettes and Round Island), and some of those visited were only partly explored. A morphologically well-characterized species of Porina (P. morelii Aptroot & Diederich) was even recently described from the Seychelles even though this archipelago can be considered as the best-explored in the south-western Indian Ocean (Diederich et al. 2017). The only species of Porina described from Réunion in the past (P. innata (Nyl.) Müll. Arg.) was not listed by van den Boom et al. (2011); indeed, it has never been reported since its description in 1858.
Future field work and subsequent taxonomic investigations are almost certain to uncover additional species of Porina on the islands of the south-western Indian Ocean. Conversely, it is likely that some epithets have been misapplied, as much of the available data is the result of using identification keys by different authors who simply chose the most similar species. Ideally, a taxonomicphylogenetic revision of Porina from these islands and of the genus Porina on a much broader geographical scale is needed to confirm those identifications. This would almost certainly change the total number of species known from the region, and, at the same time, the degree of overlap in the composition of Porina taxa between the various islands and island groups.
While species of Porinaceae are sometimes found on more than one substratum type (bark, leaves, and rock;e.g. McCarthy 2001;Lücking 2008), most Mauritian species appear to be confined to one. Among the 23 species of Porina s.l. present in Mauritius, 12 are strictly corticolous, seven strictly foliicolous, and two species occur only on rock. Two species, P. griffithsii and P. cf. eminentior, occur on bark and also on mossy rock. However, they do not appear to grow directly on the rock surface. Foliicolous species were found mainly in upland humid forests above 500 metres, except P. perminuta, which was collected in a park at 280 metres. No species were found in coastal habitats.
Apart from the new species, the most surprising discovery of the present study is Pseudosagedia crocynioides, which was previously known only from Florida (Harris 1995(Harris , 2005. In the absence of molecular data from the American populations, we cannot exclude the possibility of there being two distinct species, but, for the moment, the similar and distinctive morphology leaves little doubt of their conspecificity. This disjunct distribution is difficult to explain and might be the result of long-distance dispersal, although a direct colonization from Florida to Mauritius (or the reverse) is hardly imaginable, in particular as the species is confined to a pristine forest in Mauritius. The species might have had a wider distribution in the tropics. However, it is also possible that the current distribution is the result of the disappearance of the species over parts of its original range due to the destruction of primary forests by human activity, and we cannot exclude the discovery of other populations in the future, as the lichen flora of the tropics is still poorly known. It is probable that long-term vicariance of these forests will promote allopatric speciation between these very isolated populations.
Pseudosagedia crocynioides is most unusual within the family Porinaceae by having a byssoid thallus, a rare type that has evolved in several unrelated lineages of lichenized fungi. The species is a further prime example of parallel evolution of byssoid thalli within genera known otherwise to have a more compact thallus, besides e.g. species formerly placed in the genus Crocynia but now subsumed within Phyllopsora (Kistenich et al. 2018), Lecanactis mollis (Stirt.) Frisch & Ertz in the Roccellaceae (Ertz et al. 2015), and Thelopsis byssoidea Diederich in the Gyalectaceae (Aptroot et al. 1997;Ertz et al. 2021). A byssoid thallus is considered to be an adaptation to constantly humid climatic conditions characteristic of a tropical forest understory (Lakatos et al. 2006;Nelsen et al. 2010). It favours gas exchange because a byssoid thallus is a more ventilated structure compared to the denser organisation of most crustose lichen thalli. In many tropical Porina species, the thallus contains large amounts of calcium oxalate crystals often forming punctiform or elongate agglomerations, that render the thallus warted or ridged. The content of these structures is hydrophobic and prevents the thallus from getting completely soaked (Hafellner & Kalb 1995). Pseudosagedia crocynioides grows on large trunks in the famous pristine relict forest of Brise Fer. It is a dense humid forest inhabited by many other corticolous lichens that have developed thalli with strategies to enhance gas exchange. These include taxa having an effuse, leproid or byssoid thallus, sometimes with hydrophobic characteristics (Crocynia, Crypthonia, Herpothallon, Lepraria finkii, Phyllopsora mauritiana) or Porina species containing large amounts of oxalate crystals, e.g. Porina africana, P. florensii, and P. griffithsii.
The mtSSU phylogenetic tree includes 19 of the 23 species occurring in Mauritius. These species are dispersed throughout the phylogenetic tree, in at least eight main and strongly supported lineages, suggesting multiple colonizations of Mauritius. A new lineage strictly including Paleotropical and mainly foliicolous species (P. epiphylloides, P. longispora, P. mauritiana, and P. mazosioides) is revealed as sister to Porina s.s. (= clade from P. heterospora to P. subepiphylla, including the type of the genus, P. nucula). Our phylogenetic results bring new insights but also challenge the recent attempts to define smaller (sub)generic entities in the family. Porina griffithsii does not cluster with the other similar species having long multiseptate ascospores tapering at one end, viz. P. dolichophora and P. exasperatula (= P. dolichophora clade sensu Sobreira et al. (2018)), but it belongs to the P. simulans-P. epiphylla clade (= Phylloporina clade sensu Sobreira et al. (2018)). Moreover, the genus Saxiloba is represented by the single species sequenced so far, but instead of forming its own rather distinct lineage as shown in the original description of this genus (Lücking et al. 2020), it is part of a strongly supported lineage with saxicolous and corticolous Porina species having crustose and non-lobate thalli (e.g. P. nuculoides and P. rupicola). This raises some doubts about the use of the ascospores and thallus type alone for defining genera in the Porinaceae. Interestingly, newly sequenced species with black perithecia cluster in two lineages of species also having black perithecia: Porina borreri and P. perminuta are part of the Trichothelium clade sensu Sobreira et al. (2018), while P. covidii and P. cf. malmei form a distantly more basal lineage sister to P. byssophila+collina (P. byssophila clade sensu Sobreira et al. (2018)). The internal generic classification of the family currently suffers from the lack of a broader sampling, but more importantly also from the lack of multigene analyses that would allow a better refinement of the generic concepts.
Much remains to be done to improve the phylogeny of the Porinaceae, a highly diverse group as highlighted by our new sequences.
Porina nucula is the only species for which a mtSSU sequence was identical to a sequence available on GenBank, while other previously published species had very divergent haplotypes as shown by the long branches in our phylogenetic tree (e.g. P. internigrans, P. leptalea). Baloch & Grube (2009) demonstrated that P. epiphylla and four other species of the P. epiphylla-group (sensu Baloch & Grube 2006) include a surprisingly large number of sympatric and highly diverged haplotypes in tropical rainforests of Costa Rica. This high diversity of cryptic genetic lineages within each species is surprising, in particular as a narrow morphological concept is already in use within this group of leaf-colonizing lichens and as it was based on the mtSSU locus known to have a low intraspecific variation (Baloch & Grube 2009 and references therein). This also calls into question the use of branch length for the distinction between taxa. The highly diverged haplotypes found in the foliicolous Mauritian material first referred to the Porina epiphylla-group based on morphology strengthen the results of Baloch & Grube (2009) and call for further molecular studies of this group in the Paleotropics.
Specimens of Porina griffithsii and P. florensii are usually devoid of isidia, but a few fertile specimens were also found with isidia. Their identification was confirmed by our molecular results, although the mtSSU sequence of the isidiate specimen of the first is quite divergent ( fig. 1) and deserves further studies using a larger sampling and a more variable locus. These results suggest that the use of the presence or absence of vegetative reproductive structures such as isidia might sometimes be misleading to distinguish taxa that are otherwise morphologically similar as shown for other groups (e.g. Otálora et al. 2017).
The rich diversity of Porina s.l. found in Mauritius highlights the importance of the fragments of original forest cover that remain after four centuries of human occupation of the islands. It should encourage the considerable efforts deployed for their conservation management work to fight alien species and for the restoration of these primary forests. The real species diversity of the genus in the region of the south-western Indian Ocean is certainly underestimated and in need of further studies. The identity of several specimens remains uncertain despite the use of molecular data that are still too limited for tropical taxa. Indeed, a taxonomicphylogenetic revision of tropical Porina at a broader geographical scale is highly desirable.