Parmotrema s . l . ( Parmeliaceae , lichenized Ascomycota ) from Serra Geral slopes in central Rio Grande do Sul State , Brazil

(Parmotrema s.l. (Parmeliaceae, lichenized Ascomycota) from Serra Geral slopes in central Rio Grande do Sul State, Brazil). A survey of the parmotremoid lichens occurring in the central area of Rio Grande do Sul State (Brazil) revealed 31 species of Parmotrema A. Massal., distributed in three groups corresponding to the former genera Canomaculina Elix & Hale, Parmotrema s. str. and Rimelia Hale & Fletcher. All species are described, illustrated, commented and special notes compare similar species as an aid to identification.


Introduction
Continuing the studies on the diversity of Parmeliaceae from Southern Brazil, focused on species found in roadsides and slopes in the central Rio Grande do Sul State (Spielmann 2005, Spielmann & Marcelli 2008), we are presenting here the most diverse group found: Parmotrema s.l.
Parmotrema was proposed a long time ago (Massalongo 1860) to delimit some species belonging to the genus Parmelia Ach., and Parmotrema perforatum (Wulfen) A. Massal.was chosen as the type.The name Parmotrema reports the perforate apothecia of this species, from the Greek parmos = cup, and trema = perforation (Feige 1998).The genus was largely neglected by the contemporary workers, maybe with the exception of Krempelhuber (Hale 1984), and the species belonging to Parmotrema were dealt with in the large genus Parmelia Ach. by Zahlbruckner (1926aZahlbruckner ( , 1930)).Vainio (1890) assembled most species at present placed in Parmotrema in his Parmelia section Amphigymnia Vainio, and his delimitation was largely followed, sometimes with slight changes.Dodge (1959), for example, proposed Parmelia subgenus Amphigymnia (Vainio) Dodge, and with this name, the species were monographically treated by Hale (1965).In this meantime, some combinations in Parmotrema were proposed by M. Choisy, as can be seen in Lamb (1963).Later, Hale (1974a) recognized Parmotrema and placed in it the species studied in his monograph of Amphigymnia (Hale 1965).Therefore, Parmotrema began to be widely used.Nevertheless, some workers (e.g.Hawksworth, Dey, Krog, and Swinscow) do not accept this delimitation (Culberson 1991).Krog & Swinscow (1983) changed their mind, yet until recently (see Purvis et al. 1992) some Europeans still do not used Parmotrema (and no one of the genera proposed by Hale).Today the generic delimitation in Parmeliaceae is very controversial, but several Hale's propositions are more widely used (see, e.g., Hawksworth et al. 2008).
Rimelia was proposed by Hale & Fletcher (1990), based on Parmelia section Hypotrachyna *Irregularis (Vainio 1890).It shares several features with Parmotrema, but differs by a combination of characters: upper surface reticulate cracked, cilia almost always present [lacking only in Rimelia ruminata (Zahlbr.)Hale & Fletcher], lower surface generally rhizinate to margins (with several exceptions), and rhizines often squarrose (Hale & Fletcher 1990).However, there are species of Parmotrema with reticular maculae (particularly those with salazinic acid) and rhizinate to the margin, as well as species of Rimelia with a wide bare zone on lobes undersurface, so that the distinction of the genera is, in many instances, difficult.Canomaculina Elix & Hale, by their turn, has species with broad or narrow lobes, ciliate, upper surface with effigurate maculae, lower surface rhizinate to the margins and dimorphic rhizines (Elix 1997).
Previously we treated Canomaculina and Rimelia separately (Spielmann & Marcelli 2008), but here they are included under Parmotrema, following the proposed circumscription of Blanco et al. (2005).Although few Parmotrema s. str.were included in this paper, P. perforatum, the type of the genus, was shown to be closely related to the others representatives.Maybe future studies, based on a higher amount of species, mainly from South America, could change this picture.
Nowadays about 350 species of Parmotrema s.l. are known (Blanco et al. 2005), although much more remain to be discovered or correctly delimited, based on modern taxonomic features.

Material and methods
Detailed description and maps of the studied area, material and methods employed and considerations on general morphology can be found in Spielmann (2005) and Spielmann & Marcelli (2008).Spot tests were performed by the use of K (potassium hydroxide), C (sodium hypochlorite) and P (paraphenylenediamine).TLC analysis followed Huneck & Yoshimura (1996), Orange et al. (2001) and Bungartz (2001).Several specimens were collected "on the roadside".So this information is not repeated in the Specimens examined.Just different data were kept.Brazilian States abbreviations follow Marcelli (1998).

Results and Discussion
In the following flora we tried to put in evidence old and modern generic concepts, both in the key and in the disposition of the species, to facilitate the identification.The figures, however, are arranged in alphabetical order.
Key to Parmotrema from Serra Geral slopes in central Rio Grande do Sul State, Brazil Group I -Canomaculina-like lichens (species with dimorphic rhizines, effigurate or punctiform maculae, and rhizinate up to the margin, belonging to Canomaculina sensu Elix 1997) This group of species is distinguished by the usually thickened, furcated and spiky cilia, maculate effigurate upper cortex, lower surface often brown and usually rhizinate up to the margin and the presence of dimorphic rhizines, the later feature being the more distinctive one.
The specimen A.A. Spielmann & L.S. Canêz 668 (SP) has the lower surface strongly veined and sometimes depressed, as well as a white variegated margin.All specimens were found growing on the roadsides, in open places.(1976).
In the specimens here examined the rhizines range from abundant and covering the entire surface, to frequent, sometimes leaving some parts with few rhizines.As to the dimorphism, it ranges from sharply evident in some specimens to hardly distinguishable in others.
In the specimen A.A. Spielmann & L.S. Canêz 1298 the rhizines apparently have different sizes, but they surely are stages of development of the same type of rhizines, i.e., they're not dimorphic.This confirms the observations of Hale (1976).(1991a).
Actually P. subcaperatum is part of a complex of species, which is being studied through the types and will be published elsewhere (Spielmann & Marcelli, unpublished data).
Group II -Parmotrema-like lichens (Species without dimorphic rhizines and rarely maculate, with a broad naked rim below, belonging to Parmotrema s. str.)Parmotrema s. str. is characterized by the generally broad and loosely adnate to adnate thalli, relatively wide lobes, lower surface often with a wide bare zone along the margins and rhizines from simple to furcated or irregularly branched, but never dimorphic (Elix 1993).
To facilitate the comparison among the species of Parmotrema s. str.found, the main morphological and chemical data were summarized in the table 2.
In some species of Parmotrema the rhizines reach the margin, mainly in the form of "rhizinal papillae" (Awasthi 1976) or "rhizines in development" (Fleig 1997).Of the species here dealt with, Parmotrema internexum shows this feature.Estrabou & Adler, Mycotaxon 66: 132. 1998  Thallus pale gray, lobate, membranaceous, loosely adnate, saxicolous, 10 cm broad.Lobes irregularly branched, laterally overlapping, 3.5-8.0mm wide, surface smooth to lightly rugose, opaque to sublustrous, becoming rugose and cracked; apical zone subconcave, margin undulate, entire to broadly crenate, plane or slightly involute; lateral marginal zone undulate, plane or ascendant and involute, margin entire to crenate.Lacinules, maculae, cilia and isidia absent.Pustules simple, short-claviform or irregular, marginal, submarginal or laminal, more abundant in the proximal areas, with a lax medulla, becoming sorediose or not, arising from wrinkles or from capitate structures, often starting as black points in the thallus.Soredia granular, coming from pustules.Medulla white.Lower surface black, lustrous, smooth or more often rugose; marginal zone pale brown or ochraceous, nude or with few rhizines, smooth to papillate or sometimes rugose, opaque to sublustrous, with a sharp limit, 2-4 mm wide.Rhizines black or concolor to the marginal zone, simple or branched, sometimes with the apex flatted or coalescing in the distal areas, 0.20-0.70× 0.05-0.10mm, dispersed in groups, frequent.Apothecia and pycnidia unknown.
Parmotrema alidactylatum is characterized by the adnate and eciliate thallus with pustules, and the presence of atranorin in the medulla (K+ yellow).Parmotrema soredioaliphaticum Estrabou & Adler is very close, but its pustules become sorediose (Estrabou & Adler 1998).This subtle difference could place P. soredioaliphaticum in the synonymy of P. alidactylatum (a hypothesis admitted by Estrabou & Adler 1998).In fact, the only specimen described here show some sorediate pustules.Additional collections from Argentina and Brazil, together with the study of the types, could contribute to this question.
We are naming here the structures of this lichen as pustules, while in the original concept Estrabou & Adler (1998) Hale (1959).
It is interesting to note that Hale (1965) and Awasthi (1976) asserted that P. austrosinense has "widely perforated" apothecia, while Krog & Swinscow (1981), Elix (1994) and Nash & Elix (2002a) stated that the apothecia vary from perforate to imperforate, unfortunately not specifying if in the same thalli or not.Species pairs in which one counterpart has perforate apothecia and the other not are, as far as we are aware, unreported in Parmotrema.If both P. austrosinense and P. andinum present perforate apothecia, then the species pair is well defined.Nevertheless, if a "hidden" species, very similar to P. austrosinense, with imperforate apothecia, in fact exist, its counterpart would be another one, for example, some of the specimens reported by Awasthi (1976) as Parmelia andina, with apothecia imperforate or perforate in the center.Moreover, there are differences in the soralia and soredia between the authors that describe perforate and imperforate apothecia.Since there are several accepted synonyms in both P. austrosinense and P. andinum, only a revisionary work could solve this question, because more than one taxon seems to be implied.
The typification of this species is also a little confused.Firstly, Hale (1965) asserted that the lectotype is located in WU, and the "isotypes" in BPI and W. Elix (1994), Fleig (1997) and Eliasaro (2001) wrote "isolectotypes" instead of isotypes.On the other hand, Fleig (1997) stated that the lectotype is in US, not BPI.As some authors (e.g.Elix 1994, Eliasaro 2001) used "fide Hale 1965", andHale (1965) sends back to his previous paper, we are here using the data from Hale (1959), who made the typification.
Parmotrema bangii is characterized by the sorediose pustules and the granular soredia originated from the cortex disintegration, together with the presence of stictic acid (medulla K+ yellow, P+ slowly orange).Parmotrema perlatum is a similar sorediate species, also with stictic acid, but in this species the soralia are linear and marginal, never forming pustules.P. madylinae Fletcher has pustules, but instead of stictic acid, it has protocetraric acid (medulla K-) and shorter ascospores, 23-25 × 14-15 µm, episporium 2 µm (Hale 1986).
Unfortunately, we were unable to find pycnidia in our specimen.However, it must be noted that Krog & Swinscow (1981) reported the conidia as being bacillar to filiform, 8-10 µm long, so different from that described by Vainio (1909).
Parmotrema eciliatum is distinguished by the ciliate lobes, presence of stictic acid (medulla K+ yellow, P+ orange), apothecia imperforate, and the absence of vegetative propagules.Parmotrema aldabrense (C.W. Dodge) Hale, reported to Africa, sometimes presents stictic acid as an accessory substance to the norstictic acid (Medulla K+ yellow → red), his main substance.However, it has perforate apothecia, maculae distinct, ascospores 13-18 × 5-7 µm (Hale 1965) and filiform conidia, 12-15 µm long (Swinscow & Krog 1988).Parmotrema preperforatum (W.L. Culb.)Hale, known in the United States, also has norstictic acid and stictic acid, together with perforate apothecia (Culberson 1973), but this is the only information we were able to find about this species, so that we don't know more differences between P. aldabrense and P. preperforatum, except that one is recorded to Africa and the other to North America.Finally, P. blanchetianum (Müll.Arg.) Kalb, has similar morphology, but produces protocetraric acid instead stictic acid.In the course of the present work we noted that P. eciliatum could also be distinguished by the cilia, usually furcated or irregularly branched (figure 37).Fleig (1997).

Figures 14, 40
Thallus pale grey, lobate, membranaceous to subcoriaceous in some areas, loosely adnate, corticicolous or saxicolous, 15.0-18.5 cm broad.Lobes irregularly branched, laterally overlapping to crowded, 4-20 mm wide, surface smooth, rugose or with the cortex disintegrating in patches, sublustrous to opaque, becoming rugose in the center; apical zone plane to concave or involute, often curled or apically straighten and, combining with the involution of the lateral margins, assuming the aspect of a "T" when the lobe is observed from the lower side (figure 40), margin entire to crenate; lateral marginal zone undulate, plane to ascendant and involute, margin entire to crenate or incise-crenate (sublacinulate).Lacinules, maculae and isidia absent.Cilia black, cylindrical or flatted, simple to dichotomous, 0.40-3.00× 0.05-0.10mm, few, present on the axillar crenae, more abundant in some areas than others.Pustules developed from wrinkles, laminal to marginal, present mainly in the apices and lobe margins, breaking up in soredia.Soralia capitate to extensive, marginal to submarginal, approaching the lamina, usually developed from pustules, or sometimes directly from the fragmentation of the surface, commonly causing the "T" aspect of the lobes; soredia from subgranular to granular, sometimes isidioid and coalescing.Medulla yellow to orange.Lower surface black, lustrous, smooth, rugose or rugose-reticulate; marginal zone brown to beige or chestnut, opaque to lustrous, 1.5-6.0mm wide, nude, limit from sharp to attenuate, smooth, papillate, rugose to strongly rugose-reticulate or with veins, usually with cracks.Rhizines black or sometimes with a whitish apex, simple to furcated or irregularly branched, 0.20-3.00× 0.02-0.10mm, few to frequent, dispersed in small groups.Apothecia and pycnidia unknown.
TLC: atranorin, alectoronic acid, α-collatolic acid, skyrin and unidentified anthraquinone of Rf 49 in solvent C. Distribution: Oceania (Louwhoff & Elix 1999) and South America (Hale 1965).In South America it is known to Brazil (Zahlbruckner 1930, Marcelli 2004) and Venezuela (Feuerer 2005).In Brazil it was recorded to MG (Hale 1965), PE (Kurokawa & Moon 1998), RS (Spielmann 2006), SC (Fleig 1997) and SP (Hale 1965, Ribeiro 1998 Parmotrema hypomiltoides is distinguished by the production of soredia, often formed in arbuscular structures (like a cauliflower), the presence of alectoronic and a-collatolic acid (medulla KC+ rose → quickly orange, UV+ greenish blue) and the orange pigment K+ red (unidentified anthraquinone) close to the soralia and lobe apices.In P. rampoddense (Nyl.)Hale the soralia are usually marginal and linear, and the orange pigment K+ red (skyrin) is present only close to the lower cortex, in the old parts.Finally, P. mellissii (C.W. Dodge) Hale develops true isidia.Fleig (1997) first proposed the new combination of Parmelia hypomiltoides in Parmotrema.At that time, the status of nomenclatural propositions in theses were still somewhat obscure in the St. Louis Code (Greuter et al. 2003) Hale (1977).
However, usually the thalli are not pycnidiate, and so the conidia type is useless in the distinction of P. indicum and P. sancti-angeli.During the present study, we realized other accessory differences between these species.P. indicum often has furcated or more branched cilia, also slightly thickened (0.05-0.15 mm wide), reminding the cilia from Canomaculina (figure 39), wide marginal soralia, together with orbicular and laminal ones in the older lobes, and a robust, usually coriaceous and adnate thallus.In P. sancti-angeli, the cilia are usually simple (rarely furcated) and thinner (0.05-0.10 mm). the marginal soralia are narrow and the orbicular, laminar ones absent.The thallus is usually membranaceous, with ascendant lobes There is also the option to verify if the medulla of one specimen has norlobaridone, present only in P. indicum (Hale 1977, Krog & Swinscow 1981, Swinscow & Krog 1998).Yet this substance was not found in our specimens.
In solvent C, the norlobaridone was "masked" inside the dot of stictic acid, but in solvent A the two dots were sharply distinct.However, the presence of this substance seems not universal, since it was not detected by Eliasaro (2001) and, according to Fleig (1997) and Eliasaro & Donha (2003), it was present only in some specimens.
In the protologue of P. mordenii, Hale (1971b) stated that the soralia could occasionally form coralloid subfatiscent structures.We found in the specimen collected some pustular, capitate soralia that remember this description, however, more material is necessary to prove if this is also a good feature to take this species apart from P. praesorediosum.
Parmotrema praesorediosum is characterized by the adnate thallus, eciliate lobos, crescentshaped soralia and medulla with caperatic acid, praesorediosic acid and protolichesterinic acid (spot tests negative).In P. mordenii the soralia are usually marginal and linear, sometimes pustular and laminal, and atranorin is present in the medulla (K+ yellow).P. dilatatum (Vainio) Hale has a different chemistry, with trace of usnic acid in the cortex and medulla with protocetraric acid and echinocarpic acid (Fleig 1997).
Eliasaro & Adler (1997) stated that P. lichexanthonicum could be the parental species of P. ultralucens.Yet, P. ultralucens has imperforate apothecia, while in P. lichexanthonicum they are perforate.According to Hale (1965), species pairs with these distinct features are not known.So maybe this is the first report.
This was the more common species found in the studied area, so commonly reported (see (Spielmann 2006) at the point to be called "weedy" (Hale 1965).A very wide morphological variation is admitted, from cylindrical or coralloid isidia to isidia becoming sorediose.Maybe more taxa are being wrongly named and a revision of this group would be very welcome.
The specimens Spielmann 113 and Spielmann 354 are maculate.This feature is rarely found in the literature (just Galloway 1985), as well as the presence of lobules, reported by Krog & Swinscow (1981) and also observed in our specimens.Fleig (1997).
The epithet "wainii" is in disagreement with the recommendation 60C.1 of the Vienna Code (McNeill et al. 2007), since the end should be "oi".Ribeiro (1998) proposed "vainioi", since from 1919 Wainio changed his last name to Vainio (Alava 1998).Nevertheless, the article 60.1 asserts that "the original spelling of a name or epithet is to be retained", and the article 60.11, together with the recommendation 60C.1, deal with just the ends of the epithets, the correct spelling therefore is "wainioi".
According to current literature, P. cetratum shows a wide range of morphological variation, maybe a consequence of too broad species concepts, clearly seen by the long list of synonyms attributed to this species (see Hale & Fletcher 1990).Galloway (1985) reported several morphological differences between specimens from coastal rocks and from wood or bark inland.Fleig (1997) Hale & Fletcher (1990).

Table 2 .
Morphological and chemical features in the species of Parmotrema s. str.found (atranorin is full present in the cortex of all species).