Small Parmeliaceae (liquenized Ascomycota) of Ilhabela State Park and nearby areas, São Paulo state, Brazil

A survey of lichens at Ilhabela State Park and nearby areas in São Sebastião Island revealed the occurrence of 12 taxa belonging to six genera of small foliose Parmeliaceae, which are characterized by usually small thalli (hardly > 10 cm in diameter) with narrow lobes or laciniae less than 1 cm wide. Comments are provided for the species registered.


Introduction
With the 16th longest coastline in the world, the Brazilian Atlantic coastline extends over nearly 7.5 thousand kilometers, comprising many geographical features, including beaches, bays, reefs, and islands (CIA 2017). The Atlantic Rainforest represents the major vegetation type found in many coastal areas, including the islands. The diversity of lichens still remains poorly investigated in most coastal formations, such as restinga woods, mangroves, and rocky shores.

Material and Methods
Specimens of small foliose Parmeliaceae were collected at Ilhabela (23°45'36"S, 45°24'36"W) State Park, São Paulo state, Brazil, during several field expeditions from 2012 to 2013. Vegetational types include mostly Atlantic Rainforest, with Restinga Wood along the Eastern Coast. Accordingly to the climate classification of Köppen-Geiger, the island is classified as Tropical Aw. The altitude ranges from sea level up to almost 1400 m. The methodology used for the collection and study of lichenized fungi is described in Fink (1905), Hale (1979Hale ( , 1987, Malcolm & Galloway (1997) and . Collected specimens were observed under an Eikonal EK3ST stereomicroscope and Motic TYPE 102M microscope and the macroscopic and microscopic characteristics of somatic and reproductive structures of the thalli were analyzed. Identification was performed by comparison of published descriptions of specimens from different locations. Chemical tests consisted in spot tests, UV light exposure (360 nm), microcrystal tests and Thin Layer Chromatography (TLC) in the following eluents: (I) toluene: ethyl acetate: acetic acid 6:4:1 v/v/v and (II) toluene: acetic acid 85:15 v/v, following the methodology described in Asahina & Shibata (1954), Walker & James (1980), White & James (1985), Huneck & Yoshimura (1996) and Orange et al. (2001). All collected specimens were deposited at SP herbarium.

Results and Discussion
Thirty specimens of small Parmeliaceae were studied. We identified ten species belonging to Bulbothrix Hale ( Due to the presence of a constantly dense ombrophilous forest, it was difficult to find thalli that were mature or not damaged by chlorosis or necrosis. Even with a high search effort, almost all thalli in the study area were found in small open areas such as clearings, posts, lookouts, roadsides, and near water courses or waterfalls. We were not able to examine and identify some of the specimens collected due to the fragmented and depleted condition of the badly damaged thalli.  Benatti et al. 3412, 3413, 3418;Cocaia track, entrance to Costabela, 15.V.2013, leg., M.N. Benatti et al. 3533. This is one of the most common species of Bulbothrix in several Brazilian vegetation types (Benatti 2012d). It is mainly recognized by its larger lobed aspect (with rounded and broad lobes, often between 1 and 5 mm wide) compared to other Bulbothrix species, its distinct bulbate cilia with simple apices, the varied (often mottled) coloration of the lower cortex and the presence of medullary norstictic acid (the latter two characteristics also common to the non-isidiate B. viatica Spielmann & Marcelli). Our specimens were poorly developed with 2 to 3 cm in diameter (with the exception of a single larger, 6 cm-wide thallus), and the presence of laciniae of often small width (only a few > 3 mm wide). The laminal ciliar bulbs were scarce to common, and most specimens showed darkly mottled lower cortices, with a predominance of dark brown and black shades. All specimens have plenty of isidia, although no apotheciate material was found. The presence of atranorin and norstictic acids were confirmed by TLC.

Species
The isidiate C. caroliniana together with the sorediate C. texana (Tuck.) Elix & Hale are among the most common Canoparmelia species found in Brazil, even in urbanized or other types of anthropized areas. Interestingly, the specimen M.N. Benatti et al. 3357 presented a more accentuate medullary reaction to spot tests (almost as reddish as expected for C. sanguinea Marcelli, Benatti & Elix) but TLC confirmed traces of perlatolic acid derivatives. Although confirmed by TLC, the perlatolic acid was not clearly evidenced due to the low concentration found. This might due to the contact with other species that are growing on the same substrate and have similar compounds which could have been leached by rainfall. The atranorin and chloroatranorin acids might have occurred at low concentrations or even been absent in our material, as they were not detected on the plates. ( Benatti et al. 3456 (p.min.p.); Bairro da Feiticeira, 23º47'S, 45º21'W, 6.XII.2012, leg., M.N. Benatti et al. 3461. This species was cited for the first time for the state of São Paulo by Jungbluth (2006), who collected specimens at several spots of Cerrado vegetation in the interior of the state. This is the first record for coastal areas. The presence of capitate and raised marginal soralia and the medullary reaction to spot tests (K+ faintly rose and KC+ purple rose) are characteristic of this species. The KC reaction for the specimen M.N. Benatti 3370 gave a more purple-reddish tone. The presence of cryptochlorophaeic acid was confirmed, although atranorin might have occurred at low concentrations as it was not detected in the TLC plate. We examined several specimens of different species from the study area that presented low concentrations of atranorin, which has a solar light filter function, and interestingly, all these specimens were collected from relatively shaded places and reacted to K spot tests, showing the presence, albeit weak, of the substance. Like the specimens found by Jungbluth (2006), our specimens did not have ascomata or pycnidia. We collected a single specimen that was growing together with a specimen of Heterodermia sp. This species is characterized by the scrobiculate thallus. Atranorin, stictic and constictic acids were confirmed by TLC and menegaziaic acid was not present. An unidentified substance with Rf close to 43 was found solely in eluent I, but it could not be determined. As the specimen of C. scorbicularis was growing together with a species of Heterodermia, the substance might have leached from the later to our specimen. The thallus did not produce apothecia, pycnidia, soredia or isidia, as in the case of several other specimens found in the study area and as the propagule-producing species C. crozalsiana (B. de Lesd. ex Harm.) Lendemer & Hodk. and C. carneopruinata (Zahlbr.) Lendemer & Hodk. Therefore, it could only be identified as Crespoa cf. scrobicularis, which has been previously reported for other areas in the state of São Paulo (Jungbluth 2006;Benatti 2014b). We found only a single specimen at the study area. This species is easy recognizable due to the presence of small lobes, axillary or marginally sparse cilia that are very short and simple, a black lower cortex and medullary salazinic acid. It is similar to Parmelinella salacinifera, from which it differs by the paler and brown lower cortex, and the scarce to absent simple cilia. This material was one of the few well-developed thalli of small Parmeliaceae found (approximately 8 cm wide), very isidiate and with a single, relatively mature apothecium. The hymenia, however, did not present asci with mature ascospores. ( This isidiate species is similar to P. cinerascens, but is promptly identified by its pale brown lower cortex and the very scarce to absent marginal or axillary simple short cilia. It also presented medullary salazinic acid, which is common for Parmelinella spp. We were not able to elucidate if all specimens identified under this name belong to a single species, although they are overall nearly identical in morphology, anatomy and chemistry to all specimens described in Benatti (2014a). The only difference found was regarding the frequency of the tiny marginal cilia, with specimens completely eciliate, with a few cilia, to some with cilia in a few parts of their marginal area (Benatti 2014a). As in the case of most species registered in the present study, we only found one specimen. The only specimen collected was fertile with plenty of mature apothecia. This species commonly present hyaline, simple, ellipsoid to subglobose ascospores that measure (8-)10-15(-16) × 8-12 µm, and does not produce neither soredia nor isidia. Its morphological and chemical varieties are discussed in detail by Jungbluth (2006), but overall the single specimen examined herein was similar to those described by this author for Cerrado areas of São Paulo state. The chromatographic analysis showed the presence of atranorin and also of an unidentified substance (with a well-marked spot) with Rf 43 in the eluent 2. According to literature and simultaneous experiments, this substance might be 2,4-di-O-metilgiroforic acid, also found by Jungbluth (2006) for P. damaziana growing on Cerrado vegetation. Besides the perceptible but evanescent medullary reaction KC+ reddish-rose, other components, even possibly giroforic acid, were at too low concentrations in the extract to be detected. This species, together with P. spumosa, are one of the most common small Parmeliaceae found at the study area. It is easily recognizable in the field due its narrow and short laciniate thallus and the isidiate upper cortex. Most specimens collected, however, consisted of very small thalli of less than 4 cm wide, except for a single one of approximately 7 cm wide. The chromatographic analysis showed a few differences between specimens. Atranorin was overall present but some specimens presented three medullary compounds while others had two substances of the gyrophoric tridepsides group. Analysis by microcrystalization did not show the presence of gyrophoric acid. We were not able to compare the chromatographic behaviour due to the absence of data in the literature, but it appears to be related to the P. horrescens complex. In this study, we found only one specimen (M.N. Benatti et al. 3368) that had mature apothecia but unfortunately, the hymenium was damaged. The specimens M.N. Benatti et al. 3500 and 3368 both showed a similar chromatographical profile. The collection of more specimens and the use of other analytical techniques such as Liquid Chromatography-Mass Spectrometry (LC-MS) are needed for a conclusive characterization of these substances. (Asahina) 2012, 23º47'S, 45º21'W, leg., M.N. Benatti, L. Moro, M. Boro & C. Mayumi 3354, 3355, 3359, 3360, 3372. This species is easily recognizable by its eciliate thallus with coarse sorediate bursting pustules, differing from similar species such as P. subfatiscens, which is also eciliate but has non sorediate pustules. The chromatographical analysis showed some differences between specimens. All specimens examined contained atranorin but most had at least three other medullary compounds (some specimens contained two compounds). The low production of some substances and the leeching of other substances from other lichens present on the same substrate could justify these discrepancies. Using microcrystalization, only one thalli clearly showed presence of gyrophoric acid. Other compounds could not be determined, as documented by Jungbluth (2006) for specimens from Cerrado. The specimens M.N. Benatti et al. 3360 and 3289 were partially different chromatographically. The specimen M.N. Benatti et al. 3360 showed presence of three tridepsides together with atranorin, whereas the specimen M.N. Benatti et al. 3289 showed only two tridepsides with atranorin. Besides this difference, we were not able to make further identifications of substances using microcrystalization.

Parmelinopsis spumosa
The substances present in the specimens of both P. minarum and P. spumosa were tridepsides derived from orcinol (e.g., Hüneck & Yoshimura 1996). However, the chromatographical profiles and microcrystalization tests did not show the presence of gyrophoric acid (although they reacted C+ and KC+ rose-reddish, as expected for this substance in spot tests), even with the eluents tested (I and II) and microcrystal preparations. Again, most specimens were quite small with a few centimetres wide and therefore, they could not be used for further tests without destroying the whole specimens. This species was the only Parmeliaceae with dichotomously branched rhizines (previously in Hypotrachyna before recent studies confirmed that a group of species comprised another genus) found at the study area. It presented relatively large laciniae (up to 5 mm wide, i.e. broader and more rotund than in other species of Parmelinopsis) and abundant isidia. Atranorin and chloroatranorin were not detected by chromatography, possibly due to their low concentrations in the specimens (as the K+ yellow reaction occurred in the spot tests). The chromatography showed also three spots (Rfs around 35, 40 and 45 in eluent I) that we were not able to identify. The chemistry variability is yet not fully comprehended (e.g., Nash et al. 2002;Louwhoff & Elix 2002;Jungbluth 2006) but this is currently the only known isidiate species with dichotomously branched rhizines that show no medullary reaction to any spot tests (K, C, KC, and P).

Remototrachyna
The specimens M.N. Benatti et al. 3531 and 3336 were slightly different chromatographically, but both contained atranorin. The spots revealed with the p-anisaldehyde: sulfuric acid indicated that the substances might belong to the class of terpenes, since fatty acids are not revealed with this substance. Microcrystalization also did not show the presence of fatty acids, including protoliquesterinic and caperatic acids. Constipatic acid is very difficult to detect with this technique. Flakus et al. (2011) reported the presence of atranorin, constipatic acid complex (±, only very rarely absent) and an unidentified pigment (±, only in one specimen). Our results are in agreement with the substances reported by Flakus et al. (2011) in Bolivian specimens. It is possible that some of the specimens collected in South America and identified as R. costaricensis could represent chemotypes, since it is still premature to confirm that they belong to different species.
Aside from our work on the genus Leptogium (Kitaura et al. 2017) there are no other inventories of lichens in the study area. Therefore, this study adds new and further information to the diversity of the local mycobiota. In addition, it represents one of the few surveys (e.g., Benatti 2012b,d) performed hitherto in areas of the Atlantic Rainforest.